{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9564251537935748,
  "eval_steps": 500,
  "global_step": 11194,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0008544087491455912,
      "grad_norm": 11.00854206085205,
      "learning_rate": 1.1061797752808989e-06,
      "loss": 2.4425,
      "step": 10
    },
    {
      "epoch": 0.0017088174982911825,
      "grad_norm": 4.686764240264893,
      "learning_rate": 2.2241573033707866e-06,
      "loss": 2.1093,
      "step": 20
    },
    {
      "epoch": 0.002563226247436774,
      "grad_norm": 2.6529407501220703,
      "learning_rate": 3.342134831460675e-06,
      "loss": 1.7471,
      "step": 30
    },
    {
      "epoch": 0.003417634996582365,
      "grad_norm": 1.363114356994629,
      "learning_rate": 4.4601123595505624e-06,
      "loss": 1.4501,
      "step": 40
    },
    {
      "epoch": 0.0042720437457279565,
      "grad_norm": 1.054634690284729,
      "learning_rate": 5.5780898876404495e-06,
      "loss": 1.205,
      "step": 50
    },
    {
      "epoch": 0.005126452494873548,
      "grad_norm": 0.7446718215942383,
      "learning_rate": 6.6960674157303374e-06,
      "loss": 1.092,
      "step": 60
    },
    {
      "epoch": 0.005980861244019139,
      "grad_norm": 0.7012097835540771,
      "learning_rate": 7.814044943820226e-06,
      "loss": 1.0106,
      "step": 70
    },
    {
      "epoch": 0.00683526999316473,
      "grad_norm": 0.6184213757514954,
      "learning_rate": 8.932022471910112e-06,
      "loss": 0.9755,
      "step": 80
    },
    {
      "epoch": 0.007689678742310321,
      "grad_norm": 0.6211470365524292,
      "learning_rate": 1.005e-05,
      "loss": 1.0031,
      "step": 90
    },
    {
      "epoch": 0.008544087491455913,
      "grad_norm": 1.4690356254577637,
      "learning_rate": 1.1167977528089888e-05,
      "loss": 0.9412,
      "step": 100
    },
    {
      "epoch": 0.009398496240601503,
      "grad_norm": 0.6179498434066772,
      "learning_rate": 1.2285955056179776e-05,
      "loss": 0.9378,
      "step": 110
    },
    {
      "epoch": 0.010252904989747095,
      "grad_norm": 0.651823103427887,
      "learning_rate": 1.3403932584269664e-05,
      "loss": 0.9391,
      "step": 120
    },
    {
      "epoch": 0.011107313738892685,
      "grad_norm": 0.6029787063598633,
      "learning_rate": 1.4521910112359552e-05,
      "loss": 0.9291,
      "step": 130
    },
    {
      "epoch": 0.011961722488038277,
      "grad_norm": 0.5876877307891846,
      "learning_rate": 1.563988764044944e-05,
      "loss": 0.9231,
      "step": 140
    },
    {
      "epoch": 0.01281613123718387,
      "grad_norm": 0.5466436147689819,
      "learning_rate": 1.6757865168539328e-05,
      "loss": 0.9285,
      "step": 150
    },
    {
      "epoch": 0.01367053998632946,
      "grad_norm": 0.5822666883468628,
      "learning_rate": 1.7875842696629214e-05,
      "loss": 0.9159,
      "step": 160
    },
    {
      "epoch": 0.014524948735475052,
      "grad_norm": 0.5883556604385376,
      "learning_rate": 1.89938202247191e-05,
      "loss": 0.911,
      "step": 170
    },
    {
      "epoch": 0.015379357484620642,
      "grad_norm": 0.6157915592193604,
      "learning_rate": 2e-05,
      "loss": 0.8931,
      "step": 180
    },
    {
      "epoch": 0.016233766233766232,
      "grad_norm": 0.64449143409729,
      "learning_rate": 2e-05,
      "loss": 0.9024,
      "step": 190
    },
    {
      "epoch": 0.017088174982911826,
      "grad_norm": 0.5318751931190491,
      "learning_rate": 2e-05,
      "loss": 0.9074,
      "step": 200
    },
    {
      "epoch": 0.017942583732057416,
      "grad_norm": 0.5434000492095947,
      "learning_rate": 2e-05,
      "loss": 0.8958,
      "step": 210
    },
    {
      "epoch": 0.018796992481203006,
      "grad_norm": 0.5382893681526184,
      "learning_rate": 2e-05,
      "loss": 0.8747,
      "step": 220
    },
    {
      "epoch": 0.0196514012303486,
      "grad_norm": 0.5032609701156616,
      "learning_rate": 2e-05,
      "loss": 0.8862,
      "step": 230
    },
    {
      "epoch": 0.02050580997949419,
      "grad_norm": 0.543346643447876,
      "learning_rate": 2e-05,
      "loss": 0.8872,
      "step": 240
    },
    {
      "epoch": 0.02136021872863978,
      "grad_norm": 0.527274489402771,
      "learning_rate": 2e-05,
      "loss": 0.8818,
      "step": 250
    },
    {
      "epoch": 0.02221462747778537,
      "grad_norm": 0.5765458345413208,
      "learning_rate": 2e-05,
      "loss": 0.8687,
      "step": 260
    },
    {
      "epoch": 0.023069036226930965,
      "grad_norm": 0.5382829904556274,
      "learning_rate": 2e-05,
      "loss": 0.8915,
      "step": 270
    },
    {
      "epoch": 0.023923444976076555,
      "grad_norm": 0.5533940196037292,
      "learning_rate": 2e-05,
      "loss": 0.8891,
      "step": 280
    },
    {
      "epoch": 0.024777853725222145,
      "grad_norm": 0.5417444705963135,
      "learning_rate": 2e-05,
      "loss": 0.8847,
      "step": 290
    },
    {
      "epoch": 0.02563226247436774,
      "grad_norm": 0.5582782626152039,
      "learning_rate": 2e-05,
      "loss": 0.9424,
      "step": 300
    },
    {
      "epoch": 0.02648667122351333,
      "grad_norm": 0.5055362582206726,
      "learning_rate": 2e-05,
      "loss": 0.8781,
      "step": 310
    },
    {
      "epoch": 0.02734107997265892,
      "grad_norm": 0.5508922934532166,
      "learning_rate": 2e-05,
      "loss": 0.8775,
      "step": 320
    },
    {
      "epoch": 0.02819548872180451,
      "grad_norm": 0.5282697677612305,
      "learning_rate": 2e-05,
      "loss": 0.8627,
      "step": 330
    },
    {
      "epoch": 0.029049897470950103,
      "grad_norm": 0.5219830870628357,
      "learning_rate": 2e-05,
      "loss": 0.8577,
      "step": 340
    },
    {
      "epoch": 0.029904306220095694,
      "grad_norm": 0.5210185050964355,
      "learning_rate": 2e-05,
      "loss": 0.9101,
      "step": 350
    },
    {
      "epoch": 0.030758714969241284,
      "grad_norm": 0.5474709272384644,
      "learning_rate": 2e-05,
      "loss": 0.8914,
      "step": 360
    },
    {
      "epoch": 0.03161312371838688,
      "grad_norm": 0.5275512337684631,
      "learning_rate": 2e-05,
      "loss": 0.879,
      "step": 370
    },
    {
      "epoch": 0.032467532467532464,
      "grad_norm": 0.5367026925086975,
      "learning_rate": 2e-05,
      "loss": 0.8771,
      "step": 380
    },
    {
      "epoch": 0.03332194121667806,
      "grad_norm": 0.5371636152267456,
      "learning_rate": 2e-05,
      "loss": 0.88,
      "step": 390
    },
    {
      "epoch": 0.03417634996582365,
      "grad_norm": 0.5706266760826111,
      "learning_rate": 2e-05,
      "loss": 0.8772,
      "step": 400
    },
    {
      "epoch": 0.03503075871496924,
      "grad_norm": 0.5139058232307434,
      "learning_rate": 2e-05,
      "loss": 0.8763,
      "step": 410
    },
    {
      "epoch": 0.03588516746411483,
      "grad_norm": 0.5124853849411011,
      "learning_rate": 2e-05,
      "loss": 0.8585,
      "step": 420
    },
    {
      "epoch": 0.036739576213260426,
      "grad_norm": 0.4997583329677582,
      "learning_rate": 2e-05,
      "loss": 0.8638,
      "step": 430
    },
    {
      "epoch": 0.03759398496240601,
      "grad_norm": 0.5006685853004456,
      "learning_rate": 2e-05,
      "loss": 0.8651,
      "step": 440
    },
    {
      "epoch": 0.03844839371155161,
      "grad_norm": 0.5022147297859192,
      "learning_rate": 2e-05,
      "loss": 0.856,
      "step": 450
    },
    {
      "epoch": 0.0393028024606972,
      "grad_norm": 0.5060321688652039,
      "learning_rate": 2e-05,
      "loss": 0.8772,
      "step": 460
    },
    {
      "epoch": 0.04015721120984279,
      "grad_norm": 0.5752074718475342,
      "learning_rate": 2e-05,
      "loss": 0.8583,
      "step": 470
    },
    {
      "epoch": 0.04101161995898838,
      "grad_norm": 0.5028130412101746,
      "learning_rate": 2e-05,
      "loss": 0.8486,
      "step": 480
    },
    {
      "epoch": 0.041866028708133975,
      "grad_norm": 0.46682220697402954,
      "learning_rate": 2e-05,
      "loss": 0.8486,
      "step": 490
    },
    {
      "epoch": 0.04272043745727956,
      "grad_norm": 0.4957294464111328,
      "learning_rate": 2e-05,
      "loss": 0.8593,
      "step": 500
    },
    {
      "epoch": 0.043574846206425155,
      "grad_norm": 0.5031411051750183,
      "learning_rate": 2e-05,
      "loss": 0.8823,
      "step": 510
    },
    {
      "epoch": 0.04442925495557074,
      "grad_norm": 0.572361171245575,
      "learning_rate": 2e-05,
      "loss": 0.8664,
      "step": 520
    },
    {
      "epoch": 0.045283663704716336,
      "grad_norm": 0.4737163484096527,
      "learning_rate": 2e-05,
      "loss": 0.8273,
      "step": 530
    },
    {
      "epoch": 0.04613807245386193,
      "grad_norm": 0.5212064981460571,
      "learning_rate": 2e-05,
      "loss": 0.8265,
      "step": 540
    },
    {
      "epoch": 0.046992481203007516,
      "grad_norm": 0.4948560297489166,
      "learning_rate": 2e-05,
      "loss": 0.8636,
      "step": 550
    },
    {
      "epoch": 0.04784688995215311,
      "grad_norm": 0.5002424716949463,
      "learning_rate": 2e-05,
      "loss": 0.8632,
      "step": 560
    },
    {
      "epoch": 0.048701298701298704,
      "grad_norm": 0.6509451270103455,
      "learning_rate": 2e-05,
      "loss": 0.8552,
      "step": 570
    },
    {
      "epoch": 0.04955570745044429,
      "grad_norm": 0.4619647264480591,
      "learning_rate": 2e-05,
      "loss": 0.852,
      "step": 580
    },
    {
      "epoch": 0.050410116199589884,
      "grad_norm": 0.5018244385719299,
      "learning_rate": 2e-05,
      "loss": 0.8336,
      "step": 590
    },
    {
      "epoch": 0.05126452494873548,
      "grad_norm": 0.47025132179260254,
      "learning_rate": 2e-05,
      "loss": 0.839,
      "step": 600
    },
    {
      "epoch": 0.052118933697881065,
      "grad_norm": 0.5157033801078796,
      "learning_rate": 2e-05,
      "loss": 0.8513,
      "step": 610
    },
    {
      "epoch": 0.05297334244702666,
      "grad_norm": 0.506414532661438,
      "learning_rate": 2e-05,
      "loss": 0.8433,
      "step": 620
    },
    {
      "epoch": 0.05382775119617225,
      "grad_norm": 0.48786959052085876,
      "learning_rate": 2e-05,
      "loss": 0.8747,
      "step": 630
    },
    {
      "epoch": 0.05468215994531784,
      "grad_norm": 0.4625915586948395,
      "learning_rate": 2e-05,
      "loss": 0.8321,
      "step": 640
    },
    {
      "epoch": 0.05553656869446343,
      "grad_norm": 0.49463289976119995,
      "learning_rate": 2e-05,
      "loss": 0.8449,
      "step": 650
    },
    {
      "epoch": 0.05639097744360902,
      "grad_norm": 0.47395265102386475,
      "learning_rate": 2e-05,
      "loss": 0.867,
      "step": 660
    },
    {
      "epoch": 0.05724538619275461,
      "grad_norm": 0.49984267354011536,
      "learning_rate": 2e-05,
      "loss": 0.8622,
      "step": 670
    },
    {
      "epoch": 0.05809979494190021,
      "grad_norm": 0.47769609093666077,
      "learning_rate": 2e-05,
      "loss": 0.8332,
      "step": 680
    },
    {
      "epoch": 0.058954203691045794,
      "grad_norm": 0.4874252676963806,
      "learning_rate": 2e-05,
      "loss": 0.8662,
      "step": 690
    },
    {
      "epoch": 0.05980861244019139,
      "grad_norm": 0.4563353955745697,
      "learning_rate": 2e-05,
      "loss": 0.848,
      "step": 700
    },
    {
      "epoch": 0.06066302118933698,
      "grad_norm": 0.461029589176178,
      "learning_rate": 2e-05,
      "loss": 0.8245,
      "step": 710
    },
    {
      "epoch": 0.06151742993848257,
      "grad_norm": 0.4584606885910034,
      "learning_rate": 2e-05,
      "loss": 0.8558,
      "step": 720
    },
    {
      "epoch": 0.06237183868762816,
      "grad_norm": 0.5101044774055481,
      "learning_rate": 2e-05,
      "loss": 0.8361,
      "step": 730
    },
    {
      "epoch": 0.06322624743677376,
      "grad_norm": 0.4820401668548584,
      "learning_rate": 2e-05,
      "loss": 0.8452,
      "step": 740
    },
    {
      "epoch": 0.06408065618591935,
      "grad_norm": 0.49986669421195984,
      "learning_rate": 2e-05,
      "loss": 0.8406,
      "step": 750
    },
    {
      "epoch": 0.06493506493506493,
      "grad_norm": 0.4644050598144531,
      "learning_rate": 2e-05,
      "loss": 0.7966,
      "step": 760
    },
    {
      "epoch": 0.06578947368421052,
      "grad_norm": 0.4689406752586365,
      "learning_rate": 2e-05,
      "loss": 0.8434,
      "step": 770
    },
    {
      "epoch": 0.06664388243335612,
      "grad_norm": 0.4809776544570923,
      "learning_rate": 2e-05,
      "loss": 0.8433,
      "step": 780
    },
    {
      "epoch": 0.06749829118250171,
      "grad_norm": 0.4512708783149719,
      "learning_rate": 2e-05,
      "loss": 0.8581,
      "step": 790
    },
    {
      "epoch": 0.0683526999316473,
      "grad_norm": 0.45067834854125977,
      "learning_rate": 2e-05,
      "loss": 0.8289,
      "step": 800
    },
    {
      "epoch": 0.0692071086807929,
      "grad_norm": 0.45861950516700745,
      "learning_rate": 2e-05,
      "loss": 0.8213,
      "step": 810
    },
    {
      "epoch": 0.07006151742993848,
      "grad_norm": 0.8090688586235046,
      "learning_rate": 2e-05,
      "loss": 0.8438,
      "step": 820
    },
    {
      "epoch": 0.07091592617908407,
      "grad_norm": 0.47155073285102844,
      "learning_rate": 2e-05,
      "loss": 0.816,
      "step": 830
    },
    {
      "epoch": 0.07177033492822966,
      "grad_norm": 0.494607150554657,
      "learning_rate": 2e-05,
      "loss": 0.8342,
      "step": 840
    },
    {
      "epoch": 0.07262474367737526,
      "grad_norm": 0.4732503592967987,
      "learning_rate": 2e-05,
      "loss": 0.8551,
      "step": 850
    },
    {
      "epoch": 0.07347915242652085,
      "grad_norm": 0.45871812105178833,
      "learning_rate": 2e-05,
      "loss": 0.8466,
      "step": 860
    },
    {
      "epoch": 0.07433356117566643,
      "grad_norm": 0.4696637690067291,
      "learning_rate": 2e-05,
      "loss": 0.8548,
      "step": 870
    },
    {
      "epoch": 0.07518796992481203,
      "grad_norm": 0.49322789907455444,
      "learning_rate": 2e-05,
      "loss": 0.8635,
      "step": 880
    },
    {
      "epoch": 0.07604237867395762,
      "grad_norm": 0.4646133780479431,
      "learning_rate": 2e-05,
      "loss": 0.8431,
      "step": 890
    },
    {
      "epoch": 0.07689678742310321,
      "grad_norm": 0.4524765908718109,
      "learning_rate": 2e-05,
      "loss": 0.8164,
      "step": 900
    },
    {
      "epoch": 0.07775119617224881,
      "grad_norm": 0.4888342022895813,
      "learning_rate": 2e-05,
      "loss": 0.8548,
      "step": 910
    },
    {
      "epoch": 0.0786056049213944,
      "grad_norm": 0.43915030360221863,
      "learning_rate": 2e-05,
      "loss": 0.812,
      "step": 920
    },
    {
      "epoch": 0.07946001367053998,
      "grad_norm": 0.4614696800708771,
      "learning_rate": 2e-05,
      "loss": 0.8156,
      "step": 930
    },
    {
      "epoch": 0.08031442241968557,
      "grad_norm": 0.46576476097106934,
      "learning_rate": 2e-05,
      "loss": 0.8391,
      "step": 940
    },
    {
      "epoch": 0.08116883116883117,
      "grad_norm": 0.44493210315704346,
      "learning_rate": 2e-05,
      "loss": 0.826,
      "step": 950
    },
    {
      "epoch": 0.08202323991797676,
      "grad_norm": 0.4456934928894043,
      "learning_rate": 2e-05,
      "loss": 0.8421,
      "step": 960
    },
    {
      "epoch": 0.08287764866712236,
      "grad_norm": 0.5433761477470398,
      "learning_rate": 2e-05,
      "loss": 0.87,
      "step": 970
    },
    {
      "epoch": 0.08373205741626795,
      "grad_norm": 0.4579722285270691,
      "learning_rate": 2e-05,
      "loss": 0.8216,
      "step": 980
    },
    {
      "epoch": 0.08458646616541353,
      "grad_norm": 0.452722430229187,
      "learning_rate": 2e-05,
      "loss": 0.8337,
      "step": 990
    },
    {
      "epoch": 0.08544087491455912,
      "grad_norm": 0.431753545999527,
      "learning_rate": 2e-05,
      "loss": 0.8508,
      "step": 1000
    },
    {
      "epoch": 0.08629528366370472,
      "grad_norm": 0.5925270915031433,
      "learning_rate": 2e-05,
      "loss": 0.8105,
      "step": 1010
    },
    {
      "epoch": 0.08714969241285031,
      "grad_norm": 0.44082966446876526,
      "learning_rate": 2e-05,
      "loss": 0.8191,
      "step": 1020
    },
    {
      "epoch": 0.0880041011619959,
      "grad_norm": 0.47487062215805054,
      "learning_rate": 2e-05,
      "loss": 0.8214,
      "step": 1030
    },
    {
      "epoch": 0.08885850991114148,
      "grad_norm": 0.49121612310409546,
      "learning_rate": 2e-05,
      "loss": 0.8482,
      "step": 1040
    },
    {
      "epoch": 0.08971291866028708,
      "grad_norm": 0.45580044388771057,
      "learning_rate": 2e-05,
      "loss": 0.8376,
      "step": 1050
    },
    {
      "epoch": 0.09056732740943267,
      "grad_norm": 0.4568115174770355,
      "learning_rate": 2e-05,
      "loss": 0.8573,
      "step": 1060
    },
    {
      "epoch": 0.09142173615857827,
      "grad_norm": 0.4428640604019165,
      "learning_rate": 2e-05,
      "loss": 0.8159,
      "step": 1070
    },
    {
      "epoch": 0.09227614490772386,
      "grad_norm": 0.4813441038131714,
      "learning_rate": 2e-05,
      "loss": 0.8493,
      "step": 1080
    },
    {
      "epoch": 0.09313055365686945,
      "grad_norm": 0.4206986129283905,
      "learning_rate": 2e-05,
      "loss": 0.8088,
      "step": 1090
    },
    {
      "epoch": 0.09398496240601503,
      "grad_norm": 0.49628499150276184,
      "learning_rate": 2e-05,
      "loss": 0.815,
      "step": 1100
    },
    {
      "epoch": 0.09483937115516063,
      "grad_norm": 0.4241103529930115,
      "learning_rate": 2e-05,
      "loss": 0.8077,
      "step": 1110
    },
    {
      "epoch": 0.09569377990430622,
      "grad_norm": 0.4558090567588806,
      "learning_rate": 2e-05,
      "loss": 0.8052,
      "step": 1120
    },
    {
      "epoch": 0.09654818865345181,
      "grad_norm": 0.4508516192436218,
      "learning_rate": 2e-05,
      "loss": 0.82,
      "step": 1130
    },
    {
      "epoch": 0.09740259740259741,
      "grad_norm": 0.4674357771873474,
      "learning_rate": 2e-05,
      "loss": 0.814,
      "step": 1140
    },
    {
      "epoch": 0.09825700615174299,
      "grad_norm": 0.46306923031806946,
      "learning_rate": 2e-05,
      "loss": 0.8521,
      "step": 1150
    },
    {
      "epoch": 0.09911141490088858,
      "grad_norm": 0.43954190611839294,
      "learning_rate": 2e-05,
      "loss": 0.813,
      "step": 1160
    },
    {
      "epoch": 0.09996582365003417,
      "grad_norm": 0.4370790421962738,
      "learning_rate": 2e-05,
      "loss": 0.8143,
      "step": 1170
    },
    {
      "epoch": 0.10082023239917977,
      "grad_norm": 0.4371121823787689,
      "learning_rate": 2e-05,
      "loss": 0.836,
      "step": 1180
    },
    {
      "epoch": 0.10167464114832536,
      "grad_norm": 0.45741012692451477,
      "learning_rate": 2e-05,
      "loss": 0.83,
      "step": 1190
    },
    {
      "epoch": 0.10252904989747096,
      "grad_norm": 0.4397066831588745,
      "learning_rate": 2e-05,
      "loss": 0.8045,
      "step": 1200
    },
    {
      "epoch": 0.10338345864661654,
      "grad_norm": 0.45700281858444214,
      "learning_rate": 2e-05,
      "loss": 0.812,
      "step": 1210
    },
    {
      "epoch": 0.10423786739576213,
      "grad_norm": 0.4776708781719208,
      "learning_rate": 2e-05,
      "loss": 0.8301,
      "step": 1220
    },
    {
      "epoch": 0.10509227614490772,
      "grad_norm": 0.4925238788127899,
      "learning_rate": 2e-05,
      "loss": 0.806,
      "step": 1230
    },
    {
      "epoch": 0.10594668489405332,
      "grad_norm": 0.4427919089794159,
      "learning_rate": 2e-05,
      "loss": 0.8161,
      "step": 1240
    },
    {
      "epoch": 0.10680109364319891,
      "grad_norm": 0.45902568101882935,
      "learning_rate": 2e-05,
      "loss": 0.8254,
      "step": 1250
    },
    {
      "epoch": 0.1076555023923445,
      "grad_norm": 0.4676574468612671,
      "learning_rate": 2e-05,
      "loss": 0.8269,
      "step": 1260
    },
    {
      "epoch": 0.10850991114149008,
      "grad_norm": 0.443413645029068,
      "learning_rate": 2e-05,
      "loss": 0.8188,
      "step": 1270
    },
    {
      "epoch": 0.10936431989063568,
      "grad_norm": 0.43236881494522095,
      "learning_rate": 2e-05,
      "loss": 0.8103,
      "step": 1280
    },
    {
      "epoch": 0.11021872863978127,
      "grad_norm": 0.4279523193836212,
      "learning_rate": 2e-05,
      "loss": 0.7924,
      "step": 1290
    },
    {
      "epoch": 0.11107313738892687,
      "grad_norm": 0.4617557227611542,
      "learning_rate": 2e-05,
      "loss": 0.825,
      "step": 1300
    },
    {
      "epoch": 0.11192754613807246,
      "grad_norm": 0.4349711537361145,
      "learning_rate": 2e-05,
      "loss": 0.8275,
      "step": 1310
    },
    {
      "epoch": 0.11278195488721804,
      "grad_norm": 0.46123847365379333,
      "learning_rate": 2e-05,
      "loss": 0.808,
      "step": 1320
    },
    {
      "epoch": 0.11363636363636363,
      "grad_norm": 0.46817174553871155,
      "learning_rate": 2e-05,
      "loss": 0.8134,
      "step": 1330
    },
    {
      "epoch": 0.11449077238550923,
      "grad_norm": 0.43027958273887634,
      "learning_rate": 2e-05,
      "loss": 0.8239,
      "step": 1340
    },
    {
      "epoch": 0.11534518113465482,
      "grad_norm": 0.4255163371562958,
      "learning_rate": 2e-05,
      "loss": 0.8354,
      "step": 1350
    },
    {
      "epoch": 0.11619958988380041,
      "grad_norm": 0.42436283826828003,
      "learning_rate": 2e-05,
      "loss": 0.8334,
      "step": 1360
    },
    {
      "epoch": 0.11705399863294601,
      "grad_norm": 0.41989049315452576,
      "learning_rate": 2e-05,
      "loss": 0.8014,
      "step": 1370
    },
    {
      "epoch": 0.11790840738209159,
      "grad_norm": 0.4478135406970978,
      "learning_rate": 2e-05,
      "loss": 0.8041,
      "step": 1380
    },
    {
      "epoch": 0.11876281613123718,
      "grad_norm": 0.45139527320861816,
      "learning_rate": 2e-05,
      "loss": 0.7863,
      "step": 1390
    },
    {
      "epoch": 0.11961722488038277,
      "grad_norm": 0.5576887726783752,
      "learning_rate": 2e-05,
      "loss": 0.8462,
      "step": 1400
    },
    {
      "epoch": 0.12047163362952837,
      "grad_norm": 0.5432232618331909,
      "learning_rate": 2e-05,
      "loss": 0.8301,
      "step": 1410
    },
    {
      "epoch": 0.12132604237867396,
      "grad_norm": 0.4304180145263672,
      "learning_rate": 2e-05,
      "loss": 0.8182,
      "step": 1420
    },
    {
      "epoch": 0.12218045112781954,
      "grad_norm": 0.4250909388065338,
      "learning_rate": 2e-05,
      "loss": 0.7976,
      "step": 1430
    },
    {
      "epoch": 0.12303485987696514,
      "grad_norm": 0.4376416802406311,
      "learning_rate": 2e-05,
      "loss": 0.8251,
      "step": 1440
    },
    {
      "epoch": 0.12388926862611073,
      "grad_norm": 0.5384215712547302,
      "learning_rate": 2e-05,
      "loss": 0.8283,
      "step": 1450
    },
    {
      "epoch": 0.12474367737525632,
      "grad_norm": 0.44483479857444763,
      "learning_rate": 2e-05,
      "loss": 0.82,
      "step": 1460
    },
    {
      "epoch": 0.1255980861244019,
      "grad_norm": 0.4774000942707062,
      "learning_rate": 2e-05,
      "loss": 0.8185,
      "step": 1470
    },
    {
      "epoch": 0.1264524948735475,
      "grad_norm": 0.45032012462615967,
      "learning_rate": 2e-05,
      "loss": 0.8084,
      "step": 1480
    },
    {
      "epoch": 0.1273069036226931,
      "grad_norm": 0.43089014291763306,
      "learning_rate": 2e-05,
      "loss": 0.8098,
      "step": 1490
    },
    {
      "epoch": 0.1281613123718387,
      "grad_norm": 0.44045063853263855,
      "learning_rate": 2e-05,
      "loss": 0.802,
      "step": 1500
    },
    {
      "epoch": 0.12901572112098428,
      "grad_norm": 0.441500723361969,
      "learning_rate": 2e-05,
      "loss": 0.8227,
      "step": 1510
    },
    {
      "epoch": 0.12987012987012986,
      "grad_norm": 0.43266189098358154,
      "learning_rate": 2e-05,
      "loss": 0.814,
      "step": 1520
    },
    {
      "epoch": 0.13072453861927547,
      "grad_norm": 0.4659614562988281,
      "learning_rate": 2e-05,
      "loss": 0.8303,
      "step": 1530
    },
    {
      "epoch": 0.13157894736842105,
      "grad_norm": 0.42092642188072205,
      "learning_rate": 2e-05,
      "loss": 0.8176,
      "step": 1540
    },
    {
      "epoch": 0.13243335611756665,
      "grad_norm": 0.44579121470451355,
      "learning_rate": 2e-05,
      "loss": 0.7871,
      "step": 1550
    },
    {
      "epoch": 0.13328776486671223,
      "grad_norm": 0.6033672094345093,
      "learning_rate": 2e-05,
      "loss": 0.8148,
      "step": 1560
    },
    {
      "epoch": 0.1341421736158578,
      "grad_norm": 0.44670525193214417,
      "learning_rate": 2e-05,
      "loss": 0.7947,
      "step": 1570
    },
    {
      "epoch": 0.13499658236500342,
      "grad_norm": 0.45696425437927246,
      "learning_rate": 2e-05,
      "loss": 0.8021,
      "step": 1580
    },
    {
      "epoch": 0.135850991114149,
      "grad_norm": 0.4262278079986572,
      "learning_rate": 2e-05,
      "loss": 0.7942,
      "step": 1590
    },
    {
      "epoch": 0.1367053998632946,
      "grad_norm": 0.4282379746437073,
      "learning_rate": 2e-05,
      "loss": 0.8398,
      "step": 1600
    },
    {
      "epoch": 0.1375598086124402,
      "grad_norm": 0.4352816343307495,
      "learning_rate": 2e-05,
      "loss": 0.7778,
      "step": 1610
    },
    {
      "epoch": 0.1384142173615858,
      "grad_norm": 0.44685494899749756,
      "learning_rate": 2e-05,
      "loss": 0.788,
      "step": 1620
    },
    {
      "epoch": 0.13926862611073137,
      "grad_norm": 0.43920254707336426,
      "learning_rate": 2e-05,
      "loss": 0.8238,
      "step": 1630
    },
    {
      "epoch": 0.14012303485987695,
      "grad_norm": 0.4376579821109772,
      "learning_rate": 2e-05,
      "loss": 0.8061,
      "step": 1640
    },
    {
      "epoch": 0.14097744360902256,
      "grad_norm": 0.43427619338035583,
      "learning_rate": 2e-05,
      "loss": 0.7982,
      "step": 1650
    },
    {
      "epoch": 0.14183185235816814,
      "grad_norm": 0.478322833776474,
      "learning_rate": 2e-05,
      "loss": 0.8192,
      "step": 1660
    },
    {
      "epoch": 0.14268626110731375,
      "grad_norm": 0.43004366755485535,
      "learning_rate": 2e-05,
      "loss": 0.8136,
      "step": 1670
    },
    {
      "epoch": 0.14354066985645933,
      "grad_norm": 0.45207637548446655,
      "learning_rate": 2e-05,
      "loss": 0.8166,
      "step": 1680
    },
    {
      "epoch": 0.1443950786056049,
      "grad_norm": 0.4198301434516907,
      "learning_rate": 2e-05,
      "loss": 0.8011,
      "step": 1690
    },
    {
      "epoch": 0.14524948735475052,
      "grad_norm": 0.4201711118221283,
      "learning_rate": 2e-05,
      "loss": 0.7908,
      "step": 1700
    },
    {
      "epoch": 0.1461038961038961,
      "grad_norm": 0.4392196536064148,
      "learning_rate": 2e-05,
      "loss": 0.7841,
      "step": 1710
    },
    {
      "epoch": 0.1469583048530417,
      "grad_norm": 0.43777990341186523,
      "learning_rate": 2e-05,
      "loss": 0.7922,
      "step": 1720
    },
    {
      "epoch": 0.14781271360218728,
      "grad_norm": 0.41786715388298035,
      "learning_rate": 2e-05,
      "loss": 0.8177,
      "step": 1730
    },
    {
      "epoch": 0.14866712235133286,
      "grad_norm": 0.45031750202178955,
      "learning_rate": 2e-05,
      "loss": 0.7949,
      "step": 1740
    },
    {
      "epoch": 0.14952153110047847,
      "grad_norm": 0.4286364018917084,
      "learning_rate": 2e-05,
      "loss": 0.7885,
      "step": 1750
    },
    {
      "epoch": 0.15037593984962405,
      "grad_norm": 0.4539773166179657,
      "learning_rate": 2e-05,
      "loss": 0.8298,
      "step": 1760
    },
    {
      "epoch": 0.15123034859876966,
      "grad_norm": 0.41946274042129517,
      "learning_rate": 2e-05,
      "loss": 0.8042,
      "step": 1770
    },
    {
      "epoch": 0.15208475734791524,
      "grad_norm": 0.44184574484825134,
      "learning_rate": 2e-05,
      "loss": 0.8141,
      "step": 1780
    },
    {
      "epoch": 0.15293916609706085,
      "grad_norm": 0.453799843788147,
      "learning_rate": 2e-05,
      "loss": 0.8356,
      "step": 1790
    },
    {
      "epoch": 0.15379357484620643,
      "grad_norm": 0.46639925241470337,
      "learning_rate": 2e-05,
      "loss": 0.7995,
      "step": 1800
    },
    {
      "epoch": 0.154647983595352,
      "grad_norm": 0.42801088094711304,
      "learning_rate": 2e-05,
      "loss": 0.7778,
      "step": 1810
    },
    {
      "epoch": 0.15550239234449761,
      "grad_norm": 0.4495405852794647,
      "learning_rate": 2e-05,
      "loss": 0.799,
      "step": 1820
    },
    {
      "epoch": 0.1563568010936432,
      "grad_norm": 0.4354593753814697,
      "learning_rate": 2e-05,
      "loss": 0.8257,
      "step": 1830
    },
    {
      "epoch": 0.1572112098427888,
      "grad_norm": 0.4067724347114563,
      "learning_rate": 2e-05,
      "loss": 0.7885,
      "step": 1840
    },
    {
      "epoch": 0.15806561859193438,
      "grad_norm": 0.4517872631549835,
      "learning_rate": 2e-05,
      "loss": 0.779,
      "step": 1850
    },
    {
      "epoch": 0.15892002734107996,
      "grad_norm": 0.4226350784301758,
      "learning_rate": 2e-05,
      "loss": 0.7816,
      "step": 1860
    },
    {
      "epoch": 0.15977443609022557,
      "grad_norm": 0.43251317739486694,
      "learning_rate": 2e-05,
      "loss": 0.7999,
      "step": 1870
    },
    {
      "epoch": 0.16062884483937115,
      "grad_norm": 0.4463696777820587,
      "learning_rate": 2e-05,
      "loss": 0.816,
      "step": 1880
    },
    {
      "epoch": 0.16148325358851676,
      "grad_norm": 0.41246628761291504,
      "learning_rate": 2e-05,
      "loss": 0.7909,
      "step": 1890
    },
    {
      "epoch": 0.16233766233766234,
      "grad_norm": 0.4619990885257721,
      "learning_rate": 2e-05,
      "loss": 0.794,
      "step": 1900
    },
    {
      "epoch": 0.16319207108680792,
      "grad_norm": 0.6987309455871582,
      "learning_rate": 2e-05,
      "loss": 0.8339,
      "step": 1910
    },
    {
      "epoch": 0.16404647983595352,
      "grad_norm": 0.4621772766113281,
      "learning_rate": 2e-05,
      "loss": 0.8005,
      "step": 1920
    },
    {
      "epoch": 0.1649008885850991,
      "grad_norm": 0.4086742401123047,
      "learning_rate": 2e-05,
      "loss": 0.7964,
      "step": 1930
    },
    {
      "epoch": 0.1657552973342447,
      "grad_norm": 0.4323413670063019,
      "learning_rate": 2e-05,
      "loss": 0.7759,
      "step": 1940
    },
    {
      "epoch": 0.1666097060833903,
      "grad_norm": 0.4169401526451111,
      "learning_rate": 2e-05,
      "loss": 0.7965,
      "step": 1950
    },
    {
      "epoch": 0.1674641148325359,
      "grad_norm": 0.44284898042678833,
      "learning_rate": 2e-05,
      "loss": 0.8008,
      "step": 1960
    },
    {
      "epoch": 0.16831852358168148,
      "grad_norm": 0.44360917806625366,
      "learning_rate": 2e-05,
      "loss": 0.8244,
      "step": 1970
    },
    {
      "epoch": 0.16917293233082706,
      "grad_norm": 0.43291980028152466,
      "learning_rate": 2e-05,
      "loss": 0.7864,
      "step": 1980
    },
    {
      "epoch": 0.17002734107997267,
      "grad_norm": 0.43684226274490356,
      "learning_rate": 2e-05,
      "loss": 0.8172,
      "step": 1990
    },
    {
      "epoch": 0.17088174982911825,
      "grad_norm": 0.4126867651939392,
      "learning_rate": 2e-05,
      "loss": 0.7949,
      "step": 2000
    },
    {
      "epoch": 0.17173615857826385,
      "grad_norm": 0.4156489968299866,
      "learning_rate": 2e-05,
      "loss": 0.818,
      "step": 2010
    },
    {
      "epoch": 0.17259056732740943,
      "grad_norm": 0.4190804362297058,
      "learning_rate": 2e-05,
      "loss": 0.7805,
      "step": 2020
    },
    {
      "epoch": 0.173444976076555,
      "grad_norm": 0.4342236816883087,
      "learning_rate": 2e-05,
      "loss": 0.7805,
      "step": 2030
    },
    {
      "epoch": 0.17429938482570062,
      "grad_norm": 0.4375854730606079,
      "learning_rate": 2e-05,
      "loss": 0.8016,
      "step": 2040
    },
    {
      "epoch": 0.1751537935748462,
      "grad_norm": 0.43170157074928284,
      "learning_rate": 2e-05,
      "loss": 0.7991,
      "step": 2050
    },
    {
      "epoch": 0.1760082023239918,
      "grad_norm": 0.4225408136844635,
      "learning_rate": 2e-05,
      "loss": 0.8297,
      "step": 2060
    },
    {
      "epoch": 0.1768626110731374,
      "grad_norm": 0.4088500142097473,
      "learning_rate": 2e-05,
      "loss": 0.8231,
      "step": 2070
    },
    {
      "epoch": 0.17771701982228297,
      "grad_norm": 0.39183762669563293,
      "learning_rate": 2e-05,
      "loss": 0.7654,
      "step": 2080
    },
    {
      "epoch": 0.17857142857142858,
      "grad_norm": 0.4299352169036865,
      "learning_rate": 2e-05,
      "loss": 0.7944,
      "step": 2090
    },
    {
      "epoch": 0.17942583732057416,
      "grad_norm": 0.41311776638031006,
      "learning_rate": 2e-05,
      "loss": 0.8065,
      "step": 2100
    },
    {
      "epoch": 0.18028024606971976,
      "grad_norm": 0.4197043180465698,
      "learning_rate": 2e-05,
      "loss": 0.7967,
      "step": 2110
    },
    {
      "epoch": 0.18113465481886534,
      "grad_norm": 0.4080440104007721,
      "learning_rate": 2e-05,
      "loss": 0.7941,
      "step": 2120
    },
    {
      "epoch": 0.18198906356801095,
      "grad_norm": 0.4393218159675598,
      "learning_rate": 2e-05,
      "loss": 0.7926,
      "step": 2130
    },
    {
      "epoch": 0.18284347231715653,
      "grad_norm": 0.4172094464302063,
      "learning_rate": 2e-05,
      "loss": 0.7831,
      "step": 2140
    },
    {
      "epoch": 0.1836978810663021,
      "grad_norm": 0.4084227979183197,
      "learning_rate": 2e-05,
      "loss": 0.7879,
      "step": 2150
    },
    {
      "epoch": 0.18455228981544772,
      "grad_norm": 0.4330205023288727,
      "learning_rate": 2e-05,
      "loss": 0.7875,
      "step": 2160
    },
    {
      "epoch": 0.1854066985645933,
      "grad_norm": 0.3870986998081207,
      "learning_rate": 2e-05,
      "loss": 0.7813,
      "step": 2170
    },
    {
      "epoch": 0.1862611073137389,
      "grad_norm": 0.4238833785057068,
      "learning_rate": 2e-05,
      "loss": 0.7915,
      "step": 2180
    },
    {
      "epoch": 0.18711551606288448,
      "grad_norm": 0.4221448004245758,
      "learning_rate": 2e-05,
      "loss": 0.8056,
      "step": 2190
    },
    {
      "epoch": 0.18796992481203006,
      "grad_norm": 0.4315853416919708,
      "learning_rate": 2e-05,
      "loss": 0.7838,
      "step": 2200
    },
    {
      "epoch": 0.18882433356117567,
      "grad_norm": 0.4521797001361847,
      "learning_rate": 2e-05,
      "loss": 0.798,
      "step": 2210
    },
    {
      "epoch": 0.18967874231032125,
      "grad_norm": 0.42917776107788086,
      "learning_rate": 2e-05,
      "loss": 0.8113,
      "step": 2220
    },
    {
      "epoch": 0.19053315105946686,
      "grad_norm": 0.4716418385505676,
      "learning_rate": 2e-05,
      "loss": 0.8062,
      "step": 2230
    },
    {
      "epoch": 0.19138755980861244,
      "grad_norm": 0.4381401836872101,
      "learning_rate": 2e-05,
      "loss": 0.7671,
      "step": 2240
    },
    {
      "epoch": 0.19224196855775802,
      "grad_norm": 0.42370718717575073,
      "learning_rate": 2e-05,
      "loss": 0.7853,
      "step": 2250
    },
    {
      "epoch": 0.19309637730690363,
      "grad_norm": 0.4564455449581146,
      "learning_rate": 2e-05,
      "loss": 0.7809,
      "step": 2260
    },
    {
      "epoch": 0.1939507860560492,
      "grad_norm": 0.4647822976112366,
      "learning_rate": 2e-05,
      "loss": 0.8156,
      "step": 2270
    },
    {
      "epoch": 0.19480519480519481,
      "grad_norm": 0.43348145484924316,
      "learning_rate": 2e-05,
      "loss": 0.7993,
      "step": 2280
    },
    {
      "epoch": 0.1956596035543404,
      "grad_norm": 0.40158313512802124,
      "learning_rate": 2e-05,
      "loss": 0.7929,
      "step": 2290
    },
    {
      "epoch": 0.19651401230348597,
      "grad_norm": 0.43731987476348877,
      "learning_rate": 2e-05,
      "loss": 0.8001,
      "step": 2300
    },
    {
      "epoch": 0.19736842105263158,
      "grad_norm": 0.42699843645095825,
      "learning_rate": 2e-05,
      "loss": 0.7947,
      "step": 2310
    },
    {
      "epoch": 0.19822282980177716,
      "grad_norm": 0.4112282693386078,
      "learning_rate": 2e-05,
      "loss": 0.7848,
      "step": 2320
    },
    {
      "epoch": 0.19907723855092277,
      "grad_norm": 0.4258609116077423,
      "learning_rate": 2e-05,
      "loss": 0.7714,
      "step": 2330
    },
    {
      "epoch": 0.19993164730006835,
      "grad_norm": 0.41939055919647217,
      "learning_rate": 2e-05,
      "loss": 0.7786,
      "step": 2340
    },
    {
      "epoch": 0.20078605604921396,
      "grad_norm": 0.42235198616981506,
      "learning_rate": 2e-05,
      "loss": 0.8058,
      "step": 2350
    },
    {
      "epoch": 0.20164046479835954,
      "grad_norm": 0.4100799262523651,
      "learning_rate": 2e-05,
      "loss": 0.7895,
      "step": 2360
    },
    {
      "epoch": 0.20249487354750512,
      "grad_norm": 0.4226270020008087,
      "learning_rate": 2e-05,
      "loss": 0.8104,
      "step": 2370
    },
    {
      "epoch": 0.20334928229665072,
      "grad_norm": 0.41681861877441406,
      "learning_rate": 2e-05,
      "loss": 0.7823,
      "step": 2380
    },
    {
      "epoch": 0.2042036910457963,
      "grad_norm": 0.41445469856262207,
      "learning_rate": 2e-05,
      "loss": 0.7792,
      "step": 2390
    },
    {
      "epoch": 0.2050580997949419,
      "grad_norm": 0.4336954951286316,
      "learning_rate": 2e-05,
      "loss": 0.7979,
      "step": 2400
    },
    {
      "epoch": 0.2059125085440875,
      "grad_norm": 0.4223636984825134,
      "learning_rate": 2e-05,
      "loss": 0.8185,
      "step": 2410
    },
    {
      "epoch": 0.20676691729323307,
      "grad_norm": 0.4034973084926605,
      "learning_rate": 2e-05,
      "loss": 0.7982,
      "step": 2420
    },
    {
      "epoch": 0.20762132604237868,
      "grad_norm": 0.42055076360702515,
      "learning_rate": 2e-05,
      "loss": 0.7701,
      "step": 2430
    },
    {
      "epoch": 0.20847573479152426,
      "grad_norm": 0.4145602285861969,
      "learning_rate": 2e-05,
      "loss": 0.7973,
      "step": 2440
    },
    {
      "epoch": 0.20933014354066987,
      "grad_norm": 0.4494640827178955,
      "learning_rate": 2e-05,
      "loss": 0.7956,
      "step": 2450
    },
    {
      "epoch": 0.21018455228981545,
      "grad_norm": 0.4173319339752197,
      "learning_rate": 2e-05,
      "loss": 0.832,
      "step": 2460
    },
    {
      "epoch": 0.21103896103896103,
      "grad_norm": 0.4216117560863495,
      "learning_rate": 2e-05,
      "loss": 0.7925,
      "step": 2470
    },
    {
      "epoch": 0.21189336978810663,
      "grad_norm": 0.9099227786064148,
      "learning_rate": 2e-05,
      "loss": 0.811,
      "step": 2480
    },
    {
      "epoch": 0.2127477785372522,
      "grad_norm": 0.41247525811195374,
      "learning_rate": 2e-05,
      "loss": 0.7879,
      "step": 2490
    },
    {
      "epoch": 0.21360218728639782,
      "grad_norm": 0.40090277791023254,
      "learning_rate": 2e-05,
      "loss": 0.7759,
      "step": 2500
    },
    {
      "epoch": 0.2144565960355434,
      "grad_norm": 0.43528005480766296,
      "learning_rate": 2e-05,
      "loss": 0.742,
      "step": 2510
    },
    {
      "epoch": 0.215311004784689,
      "grad_norm": 0.40456560254096985,
      "learning_rate": 2e-05,
      "loss": 0.7851,
      "step": 2520
    },
    {
      "epoch": 0.2161654135338346,
      "grad_norm": 0.41995969414711,
      "learning_rate": 2e-05,
      "loss": 0.7725,
      "step": 2530
    },
    {
      "epoch": 0.21701982228298017,
      "grad_norm": 0.43686243891716003,
      "learning_rate": 2e-05,
      "loss": 0.7767,
      "step": 2540
    },
    {
      "epoch": 0.21787423103212578,
      "grad_norm": 0.4204292893409729,
      "learning_rate": 2e-05,
      "loss": 0.8399,
      "step": 2550
    },
    {
      "epoch": 0.21872863978127136,
      "grad_norm": 0.4287063777446747,
      "learning_rate": 2e-05,
      "loss": 0.7955,
      "step": 2560
    },
    {
      "epoch": 0.21958304853041696,
      "grad_norm": 0.4153030812740326,
      "learning_rate": 2e-05,
      "loss": 0.7573,
      "step": 2570
    },
    {
      "epoch": 0.22043745727956254,
      "grad_norm": 0.43453824520111084,
      "learning_rate": 2e-05,
      "loss": 0.7872,
      "step": 2580
    },
    {
      "epoch": 0.22129186602870812,
      "grad_norm": 0.42367810010910034,
      "learning_rate": 2e-05,
      "loss": 0.7772,
      "step": 2590
    },
    {
      "epoch": 0.22214627477785373,
      "grad_norm": 0.3947020173072815,
      "learning_rate": 2e-05,
      "loss": 0.7789,
      "step": 2600
    },
    {
      "epoch": 0.2230006835269993,
      "grad_norm": 0.39880385994911194,
      "learning_rate": 2e-05,
      "loss": 0.801,
      "step": 2610
    },
    {
      "epoch": 0.22385509227614492,
      "grad_norm": 0.41851526498794556,
      "learning_rate": 2e-05,
      "loss": 0.7848,
      "step": 2620
    },
    {
      "epoch": 0.2247095010252905,
      "grad_norm": 0.49217864871025085,
      "learning_rate": 2e-05,
      "loss": 0.7955,
      "step": 2630
    },
    {
      "epoch": 0.22556390977443608,
      "grad_norm": 0.45744040608406067,
      "learning_rate": 2e-05,
      "loss": 0.7801,
      "step": 2640
    },
    {
      "epoch": 0.22641831852358169,
      "grad_norm": 0.41490212082862854,
      "learning_rate": 2e-05,
      "loss": 0.7772,
      "step": 2650
    },
    {
      "epoch": 0.22727272727272727,
      "grad_norm": 0.4131002128124237,
      "learning_rate": 2e-05,
      "loss": 0.8039,
      "step": 2660
    },
    {
      "epoch": 0.22812713602187287,
      "grad_norm": 0.43832796812057495,
      "learning_rate": 2e-05,
      "loss": 0.8124,
      "step": 2670
    },
    {
      "epoch": 0.22898154477101845,
      "grad_norm": 0.432599812746048,
      "learning_rate": 2e-05,
      "loss": 0.8155,
      "step": 2680
    },
    {
      "epoch": 0.22983595352016406,
      "grad_norm": 0.4112294912338257,
      "learning_rate": 2e-05,
      "loss": 0.8077,
      "step": 2690
    },
    {
      "epoch": 0.23069036226930964,
      "grad_norm": 0.4199526607990265,
      "learning_rate": 2e-05,
      "loss": 0.7806,
      "step": 2700
    },
    {
      "epoch": 0.23154477101845522,
      "grad_norm": 0.45168495178222656,
      "learning_rate": 2e-05,
      "loss": 0.7692,
      "step": 2710
    },
    {
      "epoch": 0.23239917976760083,
      "grad_norm": 0.40396353602409363,
      "learning_rate": 2e-05,
      "loss": 0.799,
      "step": 2720
    },
    {
      "epoch": 0.2332535885167464,
      "grad_norm": 0.41315266489982605,
      "learning_rate": 2e-05,
      "loss": 0.7728,
      "step": 2730
    },
    {
      "epoch": 0.23410799726589201,
      "grad_norm": 0.41976848244667053,
      "learning_rate": 2e-05,
      "loss": 0.7715,
      "step": 2740
    },
    {
      "epoch": 0.2349624060150376,
      "grad_norm": 0.4033108949661255,
      "learning_rate": 2e-05,
      "loss": 0.7765,
      "step": 2750
    },
    {
      "epoch": 0.23581681476418317,
      "grad_norm": 0.4449901580810547,
      "learning_rate": 2e-05,
      "loss": 0.7929,
      "step": 2760
    },
    {
      "epoch": 0.23667122351332878,
      "grad_norm": 0.4080033302307129,
      "learning_rate": 2e-05,
      "loss": 0.7744,
      "step": 2770
    },
    {
      "epoch": 0.23752563226247436,
      "grad_norm": 0.40643179416656494,
      "learning_rate": 2e-05,
      "loss": 0.7771,
      "step": 2780
    },
    {
      "epoch": 0.23838004101161997,
      "grad_norm": 0.41164934635162354,
      "learning_rate": 2e-05,
      "loss": 0.8,
      "step": 2790
    },
    {
      "epoch": 0.23923444976076555,
      "grad_norm": 0.447868674993515,
      "learning_rate": 2e-05,
      "loss": 0.8174,
      "step": 2800
    },
    {
      "epoch": 0.24008885850991113,
      "grad_norm": 0.39664238691329956,
      "learning_rate": 2e-05,
      "loss": 0.7784,
      "step": 2810
    },
    {
      "epoch": 0.24094326725905674,
      "grad_norm": 0.4246390759944916,
      "learning_rate": 2e-05,
      "loss": 0.7671,
      "step": 2820
    },
    {
      "epoch": 0.24179767600820232,
      "grad_norm": 0.3972958028316498,
      "learning_rate": 2e-05,
      "loss": 0.7902,
      "step": 2830
    },
    {
      "epoch": 0.24265208475734792,
      "grad_norm": 0.4145045578479767,
      "learning_rate": 2e-05,
      "loss": 0.7589,
      "step": 2840
    },
    {
      "epoch": 0.2435064935064935,
      "grad_norm": 0.39986804127693176,
      "learning_rate": 2e-05,
      "loss": 0.7951,
      "step": 2850
    },
    {
      "epoch": 0.24436090225563908,
      "grad_norm": 0.4382905066013336,
      "learning_rate": 2e-05,
      "loss": 0.7945,
      "step": 2860
    },
    {
      "epoch": 0.2452153110047847,
      "grad_norm": 0.4199586808681488,
      "learning_rate": 2e-05,
      "loss": 0.7943,
      "step": 2870
    },
    {
      "epoch": 0.24606971975393027,
      "grad_norm": 0.6010065078735352,
      "learning_rate": 2e-05,
      "loss": 0.7771,
      "step": 2880
    },
    {
      "epoch": 0.24692412850307588,
      "grad_norm": 0.39158105850219727,
      "learning_rate": 2e-05,
      "loss": 0.7849,
      "step": 2890
    },
    {
      "epoch": 0.24777853725222146,
      "grad_norm": 0.4009363651275635,
      "learning_rate": 2e-05,
      "loss": 0.765,
      "step": 2900
    },
    {
      "epoch": 0.24863294600136707,
      "grad_norm": 0.4049187898635864,
      "learning_rate": 2e-05,
      "loss": 0.8042,
      "step": 2910
    },
    {
      "epoch": 0.24948735475051265,
      "grad_norm": 0.4128693640232086,
      "learning_rate": 2e-05,
      "loss": 0.8009,
      "step": 2920
    },
    {
      "epoch": 0.25034176349965825,
      "grad_norm": 0.39886823296546936,
      "learning_rate": 2e-05,
      "loss": 0.7797,
      "step": 2930
    },
    {
      "epoch": 0.2511961722488038,
      "grad_norm": 0.42408713698387146,
      "learning_rate": 2e-05,
      "loss": 0.786,
      "step": 2940
    },
    {
      "epoch": 0.2520505809979494,
      "grad_norm": 0.3976152837276459,
      "learning_rate": 2e-05,
      "loss": 0.7962,
      "step": 2950
    },
    {
      "epoch": 0.252904989747095,
      "grad_norm": 0.4111250638961792,
      "learning_rate": 2e-05,
      "loss": 0.7741,
      "step": 2960
    },
    {
      "epoch": 0.25375939849624063,
      "grad_norm": 0.39206910133361816,
      "learning_rate": 2e-05,
      "loss": 0.7619,
      "step": 2970
    },
    {
      "epoch": 0.2546138072453862,
      "grad_norm": 0.41485005617141724,
      "learning_rate": 2e-05,
      "loss": 0.7976,
      "step": 2980
    },
    {
      "epoch": 0.2554682159945318,
      "grad_norm": 0.4331391751766205,
      "learning_rate": 2e-05,
      "loss": 0.7777,
      "step": 2990
    },
    {
      "epoch": 0.2563226247436774,
      "grad_norm": 0.4180612862110138,
      "learning_rate": 2e-05,
      "loss": 0.7743,
      "step": 3000
    },
    {
      "epoch": 0.25717703349282295,
      "grad_norm": 0.4284307062625885,
      "learning_rate": 2e-05,
      "loss": 0.7744,
      "step": 3010
    },
    {
      "epoch": 0.25803144224196856,
      "grad_norm": 0.40192970633506775,
      "learning_rate": 2e-05,
      "loss": 0.749,
      "step": 3020
    },
    {
      "epoch": 0.25888585099111416,
      "grad_norm": 0.42463406920433044,
      "learning_rate": 2e-05,
      "loss": 0.7861,
      "step": 3030
    },
    {
      "epoch": 0.2597402597402597,
      "grad_norm": 0.3968360126018524,
      "learning_rate": 2e-05,
      "loss": 0.7678,
      "step": 3040
    },
    {
      "epoch": 0.2605946684894053,
      "grad_norm": 0.4153085947036743,
      "learning_rate": 2e-05,
      "loss": 0.7454,
      "step": 3050
    },
    {
      "epoch": 0.26144907723855093,
      "grad_norm": 0.4490017294883728,
      "learning_rate": 2e-05,
      "loss": 0.7871,
      "step": 3060
    },
    {
      "epoch": 0.26230348598769654,
      "grad_norm": 0.4165320098400116,
      "learning_rate": 2e-05,
      "loss": 0.7662,
      "step": 3070
    },
    {
      "epoch": 0.2631578947368421,
      "grad_norm": 0.38397857546806335,
      "learning_rate": 2e-05,
      "loss": 0.7699,
      "step": 3080
    },
    {
      "epoch": 0.2640123034859877,
      "grad_norm": 0.39957764744758606,
      "learning_rate": 2e-05,
      "loss": 0.772,
      "step": 3090
    },
    {
      "epoch": 0.2648667122351333,
      "grad_norm": 0.4143160879611969,
      "learning_rate": 2e-05,
      "loss": 0.7798,
      "step": 3100
    },
    {
      "epoch": 0.26572112098427886,
      "grad_norm": 0.3915466070175171,
      "learning_rate": 2e-05,
      "loss": 0.7539,
      "step": 3110
    },
    {
      "epoch": 0.26657552973342447,
      "grad_norm": 0.4126955568790436,
      "learning_rate": 2e-05,
      "loss": 0.778,
      "step": 3120
    },
    {
      "epoch": 0.2674299384825701,
      "grad_norm": 0.39375099539756775,
      "learning_rate": 2e-05,
      "loss": 0.7826,
      "step": 3130
    },
    {
      "epoch": 0.2682843472317156,
      "grad_norm": 0.421099454164505,
      "learning_rate": 2e-05,
      "loss": 0.752,
      "step": 3140
    },
    {
      "epoch": 0.26913875598086123,
      "grad_norm": 0.4043879508972168,
      "learning_rate": 2e-05,
      "loss": 0.7853,
      "step": 3150
    },
    {
      "epoch": 0.26999316473000684,
      "grad_norm": 0.4344448149204254,
      "learning_rate": 2e-05,
      "loss": 0.7788,
      "step": 3160
    },
    {
      "epoch": 0.27084757347915245,
      "grad_norm": 0.40575578808784485,
      "learning_rate": 2e-05,
      "loss": 0.7801,
      "step": 3170
    },
    {
      "epoch": 0.271701982228298,
      "grad_norm": 0.40416884422302246,
      "learning_rate": 2e-05,
      "loss": 0.8032,
      "step": 3180
    },
    {
      "epoch": 0.2725563909774436,
      "grad_norm": 0.4253685176372528,
      "learning_rate": 2e-05,
      "loss": 0.813,
      "step": 3190
    },
    {
      "epoch": 0.2734107997265892,
      "grad_norm": 0.3963514268398285,
      "learning_rate": 2e-05,
      "loss": 0.7774,
      "step": 3200
    },
    {
      "epoch": 0.27426520847573477,
      "grad_norm": 0.4417586624622345,
      "learning_rate": 2e-05,
      "loss": 0.7773,
      "step": 3210
    },
    {
      "epoch": 0.2751196172248804,
      "grad_norm": 0.39846548438072205,
      "learning_rate": 2e-05,
      "loss": 0.7927,
      "step": 3220
    },
    {
      "epoch": 0.275974025974026,
      "grad_norm": 0.40777456760406494,
      "learning_rate": 2e-05,
      "loss": 0.7357,
      "step": 3230
    },
    {
      "epoch": 0.2768284347231716,
      "grad_norm": 0.3965834081172943,
      "learning_rate": 2e-05,
      "loss": 0.7803,
      "step": 3240
    },
    {
      "epoch": 0.27768284347231714,
      "grad_norm": 0.41601282358169556,
      "learning_rate": 2e-05,
      "loss": 0.7635,
      "step": 3250
    },
    {
      "epoch": 0.27853725222146275,
      "grad_norm": 0.39050528407096863,
      "learning_rate": 2e-05,
      "loss": 0.7691,
      "step": 3260
    },
    {
      "epoch": 0.27939166097060836,
      "grad_norm": 0.5213533639907837,
      "learning_rate": 2e-05,
      "loss": 0.78,
      "step": 3270
    },
    {
      "epoch": 0.2802460697197539,
      "grad_norm": 0.43881452083587646,
      "learning_rate": 2e-05,
      "loss": 0.8004,
      "step": 3280
    },
    {
      "epoch": 0.2811004784688995,
      "grad_norm": 0.42476263642311096,
      "learning_rate": 2e-05,
      "loss": 0.799,
      "step": 3290
    },
    {
      "epoch": 0.2819548872180451,
      "grad_norm": 0.399971604347229,
      "learning_rate": 2e-05,
      "loss": 0.7809,
      "step": 3300
    },
    {
      "epoch": 0.2828092959671907,
      "grad_norm": 0.3991090953350067,
      "learning_rate": 2e-05,
      "loss": 0.7819,
      "step": 3310
    },
    {
      "epoch": 0.2836637047163363,
      "grad_norm": 0.4399102032184601,
      "learning_rate": 2e-05,
      "loss": 0.7689,
      "step": 3320
    },
    {
      "epoch": 0.2845181134654819,
      "grad_norm": 0.4094748795032501,
      "learning_rate": 2e-05,
      "loss": 0.773,
      "step": 3330
    },
    {
      "epoch": 0.2853725222146275,
      "grad_norm": 0.4769653081893921,
      "learning_rate": 2e-05,
      "loss": 0.7771,
      "step": 3340
    },
    {
      "epoch": 0.28622693096377305,
      "grad_norm": 0.39396145939826965,
      "learning_rate": 2e-05,
      "loss": 0.7676,
      "step": 3350
    },
    {
      "epoch": 0.28708133971291866,
      "grad_norm": 0.47275757789611816,
      "learning_rate": 2e-05,
      "loss": 0.8053,
      "step": 3360
    },
    {
      "epoch": 0.28793574846206427,
      "grad_norm": 0.4360111653804779,
      "learning_rate": 2e-05,
      "loss": 0.7803,
      "step": 3370
    },
    {
      "epoch": 0.2887901572112098,
      "grad_norm": 0.4465491771697998,
      "learning_rate": 2e-05,
      "loss": 0.7902,
      "step": 3380
    },
    {
      "epoch": 0.2896445659603554,
      "grad_norm": 0.4282015562057495,
      "learning_rate": 2e-05,
      "loss": 0.76,
      "step": 3390
    },
    {
      "epoch": 0.29049897470950103,
      "grad_norm": 0.4026396572589874,
      "learning_rate": 2e-05,
      "loss": 0.7732,
      "step": 3400
    },
    {
      "epoch": 0.29135338345864664,
      "grad_norm": 0.4072575867176056,
      "learning_rate": 2e-05,
      "loss": 0.7583,
      "step": 3410
    },
    {
      "epoch": 0.2922077922077922,
      "grad_norm": 0.3923650085926056,
      "learning_rate": 2e-05,
      "loss": 0.7572,
      "step": 3420
    },
    {
      "epoch": 0.2930622009569378,
      "grad_norm": 0.37525373697280884,
      "learning_rate": 2e-05,
      "loss": 0.7609,
      "step": 3430
    },
    {
      "epoch": 0.2939166097060834,
      "grad_norm": 0.4097144603729248,
      "learning_rate": 2e-05,
      "loss": 0.7752,
      "step": 3440
    },
    {
      "epoch": 0.29477101845522896,
      "grad_norm": 0.4037517309188843,
      "learning_rate": 2e-05,
      "loss": 0.797,
      "step": 3450
    },
    {
      "epoch": 0.29562542720437457,
      "grad_norm": 0.39452654123306274,
      "learning_rate": 2e-05,
      "loss": 0.7549,
      "step": 3460
    },
    {
      "epoch": 0.2964798359535202,
      "grad_norm": 0.46122249960899353,
      "learning_rate": 2e-05,
      "loss": 0.8029,
      "step": 3470
    },
    {
      "epoch": 0.29733424470266573,
      "grad_norm": 0.38137784600257874,
      "learning_rate": 2e-05,
      "loss": 0.7395,
      "step": 3480
    },
    {
      "epoch": 0.29818865345181134,
      "grad_norm": 0.4483271837234497,
      "learning_rate": 2e-05,
      "loss": 0.8225,
      "step": 3490
    },
    {
      "epoch": 0.29904306220095694,
      "grad_norm": 0.4055141806602478,
      "learning_rate": 2e-05,
      "loss": 0.7378,
      "step": 3500
    },
    {
      "epoch": 0.29989747095010255,
      "grad_norm": 0.43829309940338135,
      "learning_rate": 2e-05,
      "loss": 0.7737,
      "step": 3510
    },
    {
      "epoch": 0.3007518796992481,
      "grad_norm": 0.3994957506656647,
      "learning_rate": 2e-05,
      "loss": 0.747,
      "step": 3520
    },
    {
      "epoch": 0.3016062884483937,
      "grad_norm": 0.4170013964176178,
      "learning_rate": 2e-05,
      "loss": 0.7602,
      "step": 3530
    },
    {
      "epoch": 0.3024606971975393,
      "grad_norm": 0.3832671046257019,
      "learning_rate": 2e-05,
      "loss": 0.7741,
      "step": 3540
    },
    {
      "epoch": 0.30331510594668487,
      "grad_norm": 0.4173208773136139,
      "learning_rate": 2e-05,
      "loss": 0.7798,
      "step": 3550
    },
    {
      "epoch": 0.3041695146958305,
      "grad_norm": 0.4196033179759979,
      "learning_rate": 2e-05,
      "loss": 0.8139,
      "step": 3560
    },
    {
      "epoch": 0.3050239234449761,
      "grad_norm": 0.39927202463150024,
      "learning_rate": 2e-05,
      "loss": 0.7526,
      "step": 3570
    },
    {
      "epoch": 0.3058783321941217,
      "grad_norm": 0.40381738543510437,
      "learning_rate": 2e-05,
      "loss": 0.7661,
      "step": 3580
    },
    {
      "epoch": 0.30673274094326725,
      "grad_norm": 0.4107179641723633,
      "learning_rate": 2e-05,
      "loss": 0.7839,
      "step": 3590
    },
    {
      "epoch": 0.30758714969241285,
      "grad_norm": 0.413437157869339,
      "learning_rate": 2e-05,
      "loss": 0.7456,
      "step": 3600
    },
    {
      "epoch": 0.30844155844155846,
      "grad_norm": 0.41528069972991943,
      "learning_rate": 2e-05,
      "loss": 0.7908,
      "step": 3610
    },
    {
      "epoch": 0.309295967190704,
      "grad_norm": 0.40742027759552,
      "learning_rate": 2e-05,
      "loss": 0.7509,
      "step": 3620
    },
    {
      "epoch": 0.3101503759398496,
      "grad_norm": 0.42556872963905334,
      "learning_rate": 2e-05,
      "loss": 0.7474,
      "step": 3630
    },
    {
      "epoch": 0.31100478468899523,
      "grad_norm": 0.4212917685508728,
      "learning_rate": 2e-05,
      "loss": 0.7723,
      "step": 3640
    },
    {
      "epoch": 0.3118591934381408,
      "grad_norm": 0.405527800321579,
      "learning_rate": 2e-05,
      "loss": 0.7633,
      "step": 3650
    },
    {
      "epoch": 0.3127136021872864,
      "grad_norm": 0.3918667733669281,
      "learning_rate": 2e-05,
      "loss": 0.794,
      "step": 3660
    },
    {
      "epoch": 0.313568010936432,
      "grad_norm": 0.4197263717651367,
      "learning_rate": 2e-05,
      "loss": 0.7697,
      "step": 3670
    },
    {
      "epoch": 0.3144224196855776,
      "grad_norm": 0.3946523666381836,
      "learning_rate": 2e-05,
      "loss": 0.7447,
      "step": 3680
    },
    {
      "epoch": 0.31527682843472316,
      "grad_norm": 0.4318191111087799,
      "learning_rate": 2e-05,
      "loss": 0.7706,
      "step": 3690
    },
    {
      "epoch": 0.31613123718386876,
      "grad_norm": 0.38150814175605774,
      "learning_rate": 2e-05,
      "loss": 0.7465,
      "step": 3700
    },
    {
      "epoch": 0.31698564593301437,
      "grad_norm": 0.4278201162815094,
      "learning_rate": 2e-05,
      "loss": 0.7689,
      "step": 3710
    },
    {
      "epoch": 0.3178400546821599,
      "grad_norm": 0.4148176610469818,
      "learning_rate": 2e-05,
      "loss": 0.7857,
      "step": 3720
    },
    {
      "epoch": 0.31869446343130553,
      "grad_norm": 0.39185822010040283,
      "learning_rate": 2e-05,
      "loss": 0.7576,
      "step": 3730
    },
    {
      "epoch": 0.31954887218045114,
      "grad_norm": 0.38302555680274963,
      "learning_rate": 2e-05,
      "loss": 0.7717,
      "step": 3740
    },
    {
      "epoch": 0.32040328092959675,
      "grad_norm": 0.4142118990421295,
      "learning_rate": 2e-05,
      "loss": 0.7814,
      "step": 3750
    },
    {
      "epoch": 0.3212576896787423,
      "grad_norm": 0.4014316201210022,
      "learning_rate": 2e-05,
      "loss": 0.7482,
      "step": 3760
    },
    {
      "epoch": 0.3221120984278879,
      "grad_norm": 0.4141567349433899,
      "learning_rate": 2e-05,
      "loss": 0.7707,
      "step": 3770
    },
    {
      "epoch": 0.3229665071770335,
      "grad_norm": 0.4212769567966461,
      "learning_rate": 2e-05,
      "loss": 0.8012,
      "step": 3780
    },
    {
      "epoch": 0.32382091592617906,
      "grad_norm": 0.3971374034881592,
      "learning_rate": 2e-05,
      "loss": 0.7733,
      "step": 3790
    },
    {
      "epoch": 0.3246753246753247,
      "grad_norm": 0.4019508957862854,
      "learning_rate": 2e-05,
      "loss": 0.7804,
      "step": 3800
    },
    {
      "epoch": 0.3255297334244703,
      "grad_norm": 0.40491899847984314,
      "learning_rate": 2e-05,
      "loss": 0.7368,
      "step": 3810
    },
    {
      "epoch": 0.32638414217361583,
      "grad_norm": 0.38166218996047974,
      "learning_rate": 2e-05,
      "loss": 0.78,
      "step": 3820
    },
    {
      "epoch": 0.32723855092276144,
      "grad_norm": 0.422429621219635,
      "learning_rate": 2e-05,
      "loss": 0.77,
      "step": 3830
    },
    {
      "epoch": 0.32809295967190705,
      "grad_norm": 0.4231925308704376,
      "learning_rate": 2e-05,
      "loss": 0.7855,
      "step": 3840
    },
    {
      "epoch": 0.32894736842105265,
      "grad_norm": 0.42817559838294983,
      "learning_rate": 2e-05,
      "loss": 0.7794,
      "step": 3850
    },
    {
      "epoch": 0.3298017771701982,
      "grad_norm": 0.38504835963249207,
      "learning_rate": 2e-05,
      "loss": 0.7562,
      "step": 3860
    },
    {
      "epoch": 0.3306561859193438,
      "grad_norm": 0.3995634913444519,
      "learning_rate": 2e-05,
      "loss": 0.772,
      "step": 3870
    },
    {
      "epoch": 0.3315105946684894,
      "grad_norm": 0.44168439507484436,
      "learning_rate": 2e-05,
      "loss": 0.7601,
      "step": 3880
    },
    {
      "epoch": 0.332365003417635,
      "grad_norm": 0.4031580984592438,
      "learning_rate": 2e-05,
      "loss": 0.7689,
      "step": 3890
    },
    {
      "epoch": 0.3332194121667806,
      "grad_norm": 0.40626052021980286,
      "learning_rate": 2e-05,
      "loss": 0.7444,
      "step": 3900
    },
    {
      "epoch": 0.3340738209159262,
      "grad_norm": 0.3823252022266388,
      "learning_rate": 2e-05,
      "loss": 0.776,
      "step": 3910
    },
    {
      "epoch": 0.3349282296650718,
      "grad_norm": 0.3728090226650238,
      "learning_rate": 2e-05,
      "loss": 0.7481,
      "step": 3920
    },
    {
      "epoch": 0.33578263841421735,
      "grad_norm": 0.3929399251937866,
      "learning_rate": 2e-05,
      "loss": 0.7718,
      "step": 3930
    },
    {
      "epoch": 0.33663704716336296,
      "grad_norm": 0.42269980907440186,
      "learning_rate": 2e-05,
      "loss": 0.7844,
      "step": 3940
    },
    {
      "epoch": 0.33749145591250856,
      "grad_norm": 0.4141886234283447,
      "learning_rate": 2e-05,
      "loss": 0.7867,
      "step": 3950
    },
    {
      "epoch": 0.3383458646616541,
      "grad_norm": 0.4118608236312866,
      "learning_rate": 2e-05,
      "loss": 0.7751,
      "step": 3960
    },
    {
      "epoch": 0.3392002734107997,
      "grad_norm": 0.4049041271209717,
      "learning_rate": 2e-05,
      "loss": 0.7861,
      "step": 3970
    },
    {
      "epoch": 0.34005468215994533,
      "grad_norm": 0.41091078519821167,
      "learning_rate": 2e-05,
      "loss": 0.7629,
      "step": 3980
    },
    {
      "epoch": 0.3409090909090909,
      "grad_norm": 0.39125552773475647,
      "learning_rate": 2e-05,
      "loss": 0.7697,
      "step": 3990
    },
    {
      "epoch": 0.3417634996582365,
      "grad_norm": 0.4104257822036743,
      "learning_rate": 2e-05,
      "loss": 0.7845,
      "step": 4000
    },
    {
      "epoch": 0.3426179084073821,
      "grad_norm": 0.3986048400402069,
      "learning_rate": 2e-05,
      "loss": 0.7619,
      "step": 4010
    },
    {
      "epoch": 0.3434723171565277,
      "grad_norm": 0.42920932173728943,
      "learning_rate": 2e-05,
      "loss": 0.7597,
      "step": 4020
    },
    {
      "epoch": 0.34432672590567326,
      "grad_norm": 0.3850744366645813,
      "learning_rate": 2e-05,
      "loss": 0.7233,
      "step": 4030
    },
    {
      "epoch": 0.34518113465481887,
      "grad_norm": 0.41486093401908875,
      "learning_rate": 2e-05,
      "loss": 0.7811,
      "step": 4040
    },
    {
      "epoch": 0.3460355434039645,
      "grad_norm": 0.40918493270874023,
      "learning_rate": 2e-05,
      "loss": 0.7733,
      "step": 4050
    },
    {
      "epoch": 0.34688995215311,
      "grad_norm": 0.5060675144195557,
      "learning_rate": 2e-05,
      "loss": 0.7638,
      "step": 4060
    },
    {
      "epoch": 0.34774436090225563,
      "grad_norm": 0.4126531481742859,
      "learning_rate": 2e-05,
      "loss": 0.7624,
      "step": 4070
    },
    {
      "epoch": 0.34859876965140124,
      "grad_norm": 1.0571564435958862,
      "learning_rate": 2e-05,
      "loss": 0.769,
      "step": 4080
    },
    {
      "epoch": 0.34945317840054685,
      "grad_norm": 0.3744198679924011,
      "learning_rate": 2e-05,
      "loss": 0.7678,
      "step": 4090
    },
    {
      "epoch": 0.3503075871496924,
      "grad_norm": 0.4054426848888397,
      "learning_rate": 2e-05,
      "loss": 0.7649,
      "step": 4100
    },
    {
      "epoch": 0.351161995898838,
      "grad_norm": 0.40937623381614685,
      "learning_rate": 2e-05,
      "loss": 0.7734,
      "step": 4110
    },
    {
      "epoch": 0.3520164046479836,
      "grad_norm": 0.4030979871749878,
      "learning_rate": 2e-05,
      "loss": 0.7792,
      "step": 4120
    },
    {
      "epoch": 0.35287081339712917,
      "grad_norm": 0.40039506554603577,
      "learning_rate": 2e-05,
      "loss": 0.7747,
      "step": 4130
    },
    {
      "epoch": 0.3537252221462748,
      "grad_norm": 0.4101240336894989,
      "learning_rate": 2e-05,
      "loss": 0.7474,
      "step": 4140
    },
    {
      "epoch": 0.3545796308954204,
      "grad_norm": 0.3959333896636963,
      "learning_rate": 2e-05,
      "loss": 0.7566,
      "step": 4150
    },
    {
      "epoch": 0.35543403964456594,
      "grad_norm": 0.4016161262989044,
      "learning_rate": 2e-05,
      "loss": 0.7862,
      "step": 4160
    },
    {
      "epoch": 0.35628844839371154,
      "grad_norm": 0.404707670211792,
      "learning_rate": 2e-05,
      "loss": 0.7766,
      "step": 4170
    },
    {
      "epoch": 0.35714285714285715,
      "grad_norm": 0.3973741829395294,
      "learning_rate": 2e-05,
      "loss": 0.7662,
      "step": 4180
    },
    {
      "epoch": 0.35799726589200276,
      "grad_norm": 0.4071394205093384,
      "learning_rate": 2e-05,
      "loss": 0.7484,
      "step": 4190
    },
    {
      "epoch": 0.3588516746411483,
      "grad_norm": 0.8778959512710571,
      "learning_rate": 2e-05,
      "loss": 0.751,
      "step": 4200
    },
    {
      "epoch": 0.3597060833902939,
      "grad_norm": 0.4312654435634613,
      "learning_rate": 2e-05,
      "loss": 0.7591,
      "step": 4210
    },
    {
      "epoch": 0.3605604921394395,
      "grad_norm": 0.39795437455177307,
      "learning_rate": 2e-05,
      "loss": 0.7498,
      "step": 4220
    },
    {
      "epoch": 0.3614149008885851,
      "grad_norm": 0.4243643879890442,
      "learning_rate": 2e-05,
      "loss": 0.777,
      "step": 4230
    },
    {
      "epoch": 0.3622693096377307,
      "grad_norm": 0.393503874540329,
      "learning_rate": 2e-05,
      "loss": 0.762,
      "step": 4240
    },
    {
      "epoch": 0.3631237183868763,
      "grad_norm": 0.4122103452682495,
      "learning_rate": 2e-05,
      "loss": 0.7737,
      "step": 4250
    },
    {
      "epoch": 0.3639781271360219,
      "grad_norm": 0.43337583541870117,
      "learning_rate": 2e-05,
      "loss": 0.7699,
      "step": 4260
    },
    {
      "epoch": 0.36483253588516745,
      "grad_norm": 0.38475391268730164,
      "learning_rate": 2e-05,
      "loss": 0.7711,
      "step": 4270
    },
    {
      "epoch": 0.36568694463431306,
      "grad_norm": 0.40655094385147095,
      "learning_rate": 2e-05,
      "loss": 0.7812,
      "step": 4280
    },
    {
      "epoch": 0.36654135338345867,
      "grad_norm": 0.39318129420280457,
      "learning_rate": 2e-05,
      "loss": 0.74,
      "step": 4290
    },
    {
      "epoch": 0.3673957621326042,
      "grad_norm": 0.39770469069480896,
      "learning_rate": 2e-05,
      "loss": 0.7746,
      "step": 4300
    },
    {
      "epoch": 0.3682501708817498,
      "grad_norm": 0.3997613489627838,
      "learning_rate": 2e-05,
      "loss": 0.7684,
      "step": 4310
    },
    {
      "epoch": 0.36910457963089544,
      "grad_norm": 0.3889233469963074,
      "learning_rate": 2e-05,
      "loss": 0.7703,
      "step": 4320
    },
    {
      "epoch": 0.369958988380041,
      "grad_norm": 0.40742871165275574,
      "learning_rate": 2e-05,
      "loss": 0.7517,
      "step": 4330
    },
    {
      "epoch": 0.3708133971291866,
      "grad_norm": 0.4064273536205292,
      "learning_rate": 2e-05,
      "loss": 0.749,
      "step": 4340
    },
    {
      "epoch": 0.3716678058783322,
      "grad_norm": 0.4193180203437805,
      "learning_rate": 2e-05,
      "loss": 0.7609,
      "step": 4350
    },
    {
      "epoch": 0.3725222146274778,
      "grad_norm": 0.4050132632255554,
      "learning_rate": 2e-05,
      "loss": 0.7693,
      "step": 4360
    },
    {
      "epoch": 0.37337662337662336,
      "grad_norm": 0.39398810267448425,
      "learning_rate": 2e-05,
      "loss": 0.7357,
      "step": 4370
    },
    {
      "epoch": 0.37423103212576897,
      "grad_norm": 0.422772079706192,
      "learning_rate": 2e-05,
      "loss": 0.7697,
      "step": 4380
    },
    {
      "epoch": 0.3750854408749146,
      "grad_norm": 0.43367302417755127,
      "learning_rate": 2e-05,
      "loss": 0.7597,
      "step": 4390
    },
    {
      "epoch": 0.37593984962406013,
      "grad_norm": 0.4362691044807434,
      "learning_rate": 2e-05,
      "loss": 0.7866,
      "step": 4400
    },
    {
      "epoch": 0.37679425837320574,
      "grad_norm": 0.38281482458114624,
      "learning_rate": 2e-05,
      "loss": 0.7399,
      "step": 4410
    },
    {
      "epoch": 0.37764866712235134,
      "grad_norm": 0.42784583568573,
      "learning_rate": 2e-05,
      "loss": 0.7846,
      "step": 4420
    },
    {
      "epoch": 0.3785030758714969,
      "grad_norm": 0.39173761010169983,
      "learning_rate": 2e-05,
      "loss": 0.7467,
      "step": 4430
    },
    {
      "epoch": 0.3793574846206425,
      "grad_norm": 0.4276053011417389,
      "learning_rate": 2e-05,
      "loss": 0.7594,
      "step": 4440
    },
    {
      "epoch": 0.3802118933697881,
      "grad_norm": 0.5029833912849426,
      "learning_rate": 2e-05,
      "loss": 0.7539,
      "step": 4450
    },
    {
      "epoch": 0.3810663021189337,
      "grad_norm": 0.4133250415325165,
      "learning_rate": 2e-05,
      "loss": 0.7607,
      "step": 4460
    },
    {
      "epoch": 0.38192071086807927,
      "grad_norm": 0.3917803466320038,
      "learning_rate": 2e-05,
      "loss": 0.736,
      "step": 4470
    },
    {
      "epoch": 0.3827751196172249,
      "grad_norm": 0.3776458203792572,
      "learning_rate": 2e-05,
      "loss": 0.7621,
      "step": 4480
    },
    {
      "epoch": 0.3836295283663705,
      "grad_norm": 0.40102991461753845,
      "learning_rate": 2e-05,
      "loss": 0.7746,
      "step": 4490
    },
    {
      "epoch": 0.38448393711551604,
      "grad_norm": 0.3935014009475708,
      "learning_rate": 2e-05,
      "loss": 0.7691,
      "step": 4500
    },
    {
      "epoch": 0.38533834586466165,
      "grad_norm": 0.38191425800323486,
      "learning_rate": 2e-05,
      "loss": 0.7568,
      "step": 4510
    },
    {
      "epoch": 0.38619275461380725,
      "grad_norm": 0.40427932143211365,
      "learning_rate": 2e-05,
      "loss": 0.7555,
      "step": 4520
    },
    {
      "epoch": 0.38704716336295286,
      "grad_norm": 0.3987480700016022,
      "learning_rate": 2e-05,
      "loss": 0.7637,
      "step": 4530
    },
    {
      "epoch": 0.3879015721120984,
      "grad_norm": 0.39704596996307373,
      "learning_rate": 2e-05,
      "loss": 0.7692,
      "step": 4540
    },
    {
      "epoch": 0.388755980861244,
      "grad_norm": 0.38936182856559753,
      "learning_rate": 2e-05,
      "loss": 0.76,
      "step": 4550
    },
    {
      "epoch": 0.38961038961038963,
      "grad_norm": 0.392251193523407,
      "learning_rate": 2e-05,
      "loss": 0.7739,
      "step": 4560
    },
    {
      "epoch": 0.3904647983595352,
      "grad_norm": 0.413920134305954,
      "learning_rate": 2e-05,
      "loss": 0.7672,
      "step": 4570
    },
    {
      "epoch": 0.3913192071086808,
      "grad_norm": 0.3814871609210968,
      "learning_rate": 2e-05,
      "loss": 0.747,
      "step": 4580
    },
    {
      "epoch": 0.3921736158578264,
      "grad_norm": 0.4402540624141693,
      "learning_rate": 2e-05,
      "loss": 0.7489,
      "step": 4590
    },
    {
      "epoch": 0.39302802460697195,
      "grad_norm": 0.45276889204978943,
      "learning_rate": 2e-05,
      "loss": 0.7493,
      "step": 4600
    },
    {
      "epoch": 0.39388243335611756,
      "grad_norm": 0.4238435626029968,
      "learning_rate": 2e-05,
      "loss": 0.754,
      "step": 4610
    },
    {
      "epoch": 0.39473684210526316,
      "grad_norm": 0.42151281237602234,
      "learning_rate": 2e-05,
      "loss": 0.7797,
      "step": 4620
    },
    {
      "epoch": 0.39559125085440877,
      "grad_norm": 0.38528621196746826,
      "learning_rate": 2e-05,
      "loss": 0.786,
      "step": 4630
    },
    {
      "epoch": 0.3964456596035543,
      "grad_norm": 0.4101194739341736,
      "learning_rate": 2e-05,
      "loss": 0.7767,
      "step": 4640
    },
    {
      "epoch": 0.39730006835269993,
      "grad_norm": 0.4006814658641815,
      "learning_rate": 2e-05,
      "loss": 0.7976,
      "step": 4650
    },
    {
      "epoch": 0.39815447710184554,
      "grad_norm": 0.4116314947605133,
      "learning_rate": 2e-05,
      "loss": 0.7525,
      "step": 4660
    },
    {
      "epoch": 0.3990088858509911,
      "grad_norm": 0.3980487883090973,
      "learning_rate": 2e-05,
      "loss": 0.7573,
      "step": 4670
    },
    {
      "epoch": 0.3998632946001367,
      "grad_norm": 0.38866686820983887,
      "learning_rate": 2e-05,
      "loss": 0.7487,
      "step": 4680
    },
    {
      "epoch": 0.4007177033492823,
      "grad_norm": 0.4014759957790375,
      "learning_rate": 2e-05,
      "loss": 0.7326,
      "step": 4690
    },
    {
      "epoch": 0.4015721120984279,
      "grad_norm": 0.403236448764801,
      "learning_rate": 2e-05,
      "loss": 0.7768,
      "step": 4700
    },
    {
      "epoch": 0.40242652084757347,
      "grad_norm": 0.3917689919471741,
      "learning_rate": 2e-05,
      "loss": 0.7246,
      "step": 4710
    },
    {
      "epoch": 0.4032809295967191,
      "grad_norm": 0.37933942675590515,
      "learning_rate": 2e-05,
      "loss": 0.76,
      "step": 4720
    },
    {
      "epoch": 0.4041353383458647,
      "grad_norm": 0.38117894530296326,
      "learning_rate": 2e-05,
      "loss": 0.7264,
      "step": 4730
    },
    {
      "epoch": 0.40498974709501023,
      "grad_norm": 0.4002052843570709,
      "learning_rate": 2e-05,
      "loss": 0.7398,
      "step": 4740
    },
    {
      "epoch": 0.40584415584415584,
      "grad_norm": 0.4108046293258667,
      "learning_rate": 2e-05,
      "loss": 0.7403,
      "step": 4750
    },
    {
      "epoch": 0.40669856459330145,
      "grad_norm": 0.3776397705078125,
      "learning_rate": 2e-05,
      "loss": 0.7707,
      "step": 4760
    },
    {
      "epoch": 0.407552973342447,
      "grad_norm": 0.3892161250114441,
      "learning_rate": 2e-05,
      "loss": 0.7654,
      "step": 4770
    },
    {
      "epoch": 0.4084073820915926,
      "grad_norm": 0.40108174085617065,
      "learning_rate": 2e-05,
      "loss": 0.7858,
      "step": 4780
    },
    {
      "epoch": 0.4092617908407382,
      "grad_norm": 0.40284615755081177,
      "learning_rate": 2e-05,
      "loss": 0.7419,
      "step": 4790
    },
    {
      "epoch": 0.4101161995898838,
      "grad_norm": 0.4094385504722595,
      "learning_rate": 2e-05,
      "loss": 0.7518,
      "step": 4800
    },
    {
      "epoch": 0.4109706083390294,
      "grad_norm": 0.382321834564209,
      "learning_rate": 2e-05,
      "loss": 0.739,
      "step": 4810
    },
    {
      "epoch": 0.411825017088175,
      "grad_norm": 0.39105379581451416,
      "learning_rate": 2e-05,
      "loss": 0.7836,
      "step": 4820
    },
    {
      "epoch": 0.4126794258373206,
      "grad_norm": 0.38638919591903687,
      "learning_rate": 2e-05,
      "loss": 0.7618,
      "step": 4830
    },
    {
      "epoch": 0.41353383458646614,
      "grad_norm": 0.42078790068626404,
      "learning_rate": 2e-05,
      "loss": 0.7587,
      "step": 4840
    },
    {
      "epoch": 0.41438824333561175,
      "grad_norm": 0.4311010539531708,
      "learning_rate": 2e-05,
      "loss": 0.7694,
      "step": 4850
    },
    {
      "epoch": 0.41524265208475736,
      "grad_norm": 0.40017813444137573,
      "learning_rate": 2e-05,
      "loss": 0.7443,
      "step": 4860
    },
    {
      "epoch": 0.41609706083390297,
      "grad_norm": 0.41783082485198975,
      "learning_rate": 2e-05,
      "loss": 0.7504,
      "step": 4870
    },
    {
      "epoch": 0.4169514695830485,
      "grad_norm": 0.3885749876499176,
      "learning_rate": 2e-05,
      "loss": 0.7798,
      "step": 4880
    },
    {
      "epoch": 0.4178058783321941,
      "grad_norm": 0.3940373957157135,
      "learning_rate": 2e-05,
      "loss": 0.7614,
      "step": 4890
    },
    {
      "epoch": 0.41866028708133973,
      "grad_norm": 0.41718438267707825,
      "learning_rate": 2e-05,
      "loss": 0.744,
      "step": 4900
    },
    {
      "epoch": 0.4195146958304853,
      "grad_norm": 0.4303945004940033,
      "learning_rate": 2e-05,
      "loss": 0.7831,
      "step": 4910
    },
    {
      "epoch": 0.4203691045796309,
      "grad_norm": 0.3983951508998871,
      "learning_rate": 2e-05,
      "loss": 0.7625,
      "step": 4920
    },
    {
      "epoch": 0.4212235133287765,
      "grad_norm": 0.39108559489250183,
      "learning_rate": 2e-05,
      "loss": 0.7602,
      "step": 4930
    },
    {
      "epoch": 0.42207792207792205,
      "grad_norm": 0.3912990093231201,
      "learning_rate": 2e-05,
      "loss": 0.7241,
      "step": 4940
    },
    {
      "epoch": 0.42293233082706766,
      "grad_norm": 0.4086664915084839,
      "learning_rate": 2e-05,
      "loss": 0.7543,
      "step": 4950
    },
    {
      "epoch": 0.42378673957621327,
      "grad_norm": 0.43198081851005554,
      "learning_rate": 2e-05,
      "loss": 0.7426,
      "step": 4960
    },
    {
      "epoch": 0.4246411483253589,
      "grad_norm": 0.37144365906715393,
      "learning_rate": 2e-05,
      "loss": 0.7376,
      "step": 4970
    },
    {
      "epoch": 0.4254955570745044,
      "grad_norm": 0.3948076367378235,
      "learning_rate": 2e-05,
      "loss": 0.7499,
      "step": 4980
    },
    {
      "epoch": 0.42634996582365003,
      "grad_norm": 0.4113522469997406,
      "learning_rate": 2e-05,
      "loss": 0.7657,
      "step": 4990
    },
    {
      "epoch": 0.42720437457279564,
      "grad_norm": 0.3880317211151123,
      "learning_rate": 2e-05,
      "loss": 0.7534,
      "step": 5000
    },
    {
      "epoch": 0.4280587833219412,
      "grad_norm": 0.3677981495857239,
      "learning_rate": 2e-05,
      "loss": 0.752,
      "step": 5010
    },
    {
      "epoch": 0.4289131920710868,
      "grad_norm": 0.4228217601776123,
      "learning_rate": 2e-05,
      "loss": 0.7637,
      "step": 5020
    },
    {
      "epoch": 0.4297676008202324,
      "grad_norm": 0.38919076323509216,
      "learning_rate": 2e-05,
      "loss": 0.7473,
      "step": 5030
    },
    {
      "epoch": 0.430622009569378,
      "grad_norm": 0.4081915616989136,
      "learning_rate": 2e-05,
      "loss": 0.7675,
      "step": 5040
    },
    {
      "epoch": 0.43147641831852357,
      "grad_norm": 0.4186146855354309,
      "learning_rate": 2e-05,
      "loss": 0.7491,
      "step": 5050
    },
    {
      "epoch": 0.4323308270676692,
      "grad_norm": 0.4249284267425537,
      "learning_rate": 2e-05,
      "loss": 0.7577,
      "step": 5060
    },
    {
      "epoch": 0.4331852358168148,
      "grad_norm": 0.5370435118675232,
      "learning_rate": 2e-05,
      "loss": 0.7321,
      "step": 5070
    },
    {
      "epoch": 0.43403964456596034,
      "grad_norm": 0.40723684430122375,
      "learning_rate": 2e-05,
      "loss": 0.7369,
      "step": 5080
    },
    {
      "epoch": 0.43489405331510594,
      "grad_norm": 0.3778870105743408,
      "learning_rate": 2e-05,
      "loss": 0.7351,
      "step": 5090
    },
    {
      "epoch": 0.43574846206425155,
      "grad_norm": 0.41734758019447327,
      "learning_rate": 2e-05,
      "loss": 0.7645,
      "step": 5100
    },
    {
      "epoch": 0.4366028708133971,
      "grad_norm": 0.42623525857925415,
      "learning_rate": 2e-05,
      "loss": 0.7746,
      "step": 5110
    },
    {
      "epoch": 0.4374572795625427,
      "grad_norm": 0.4156603515148163,
      "learning_rate": 2e-05,
      "loss": 0.7267,
      "step": 5120
    },
    {
      "epoch": 0.4383116883116883,
      "grad_norm": 0.4196552336215973,
      "learning_rate": 2e-05,
      "loss": 0.7358,
      "step": 5130
    },
    {
      "epoch": 0.4391660970608339,
      "grad_norm": 0.4046937823295593,
      "learning_rate": 2e-05,
      "loss": 0.7455,
      "step": 5140
    },
    {
      "epoch": 0.4400205058099795,
      "grad_norm": 0.42020198702812195,
      "learning_rate": 2e-05,
      "loss": 0.7626,
      "step": 5150
    },
    {
      "epoch": 0.4408749145591251,
      "grad_norm": 0.40685462951660156,
      "learning_rate": 2e-05,
      "loss": 0.7422,
      "step": 5160
    },
    {
      "epoch": 0.4417293233082707,
      "grad_norm": 0.4485331177711487,
      "learning_rate": 2e-05,
      "loss": 0.7545,
      "step": 5170
    },
    {
      "epoch": 0.44258373205741625,
      "grad_norm": 0.39519354701042175,
      "learning_rate": 2e-05,
      "loss": 0.7351,
      "step": 5180
    },
    {
      "epoch": 0.44343814080656185,
      "grad_norm": 0.4375475347042084,
      "learning_rate": 2e-05,
      "loss": 0.7681,
      "step": 5190
    },
    {
      "epoch": 0.44429254955570746,
      "grad_norm": 0.4351958930492401,
      "learning_rate": 2e-05,
      "loss": 0.759,
      "step": 5200
    },
    {
      "epoch": 0.44514695830485307,
      "grad_norm": 0.4109622538089752,
      "learning_rate": 2e-05,
      "loss": 0.7606,
      "step": 5210
    },
    {
      "epoch": 0.4460013670539986,
      "grad_norm": 0.4069000482559204,
      "learning_rate": 2e-05,
      "loss": 0.7715,
      "step": 5220
    },
    {
      "epoch": 0.44685577580314423,
      "grad_norm": 0.3997276723384857,
      "learning_rate": 2e-05,
      "loss": 0.7575,
      "step": 5230
    },
    {
      "epoch": 0.44771018455228984,
      "grad_norm": 0.4074077308177948,
      "learning_rate": 2e-05,
      "loss": 0.7459,
      "step": 5240
    },
    {
      "epoch": 0.4485645933014354,
      "grad_norm": 0.3978741466999054,
      "learning_rate": 2e-05,
      "loss": 0.7493,
      "step": 5250
    },
    {
      "epoch": 0.449419002050581,
      "grad_norm": 0.39086443185806274,
      "learning_rate": 2e-05,
      "loss": 0.7415,
      "step": 5260
    },
    {
      "epoch": 0.4502734107997266,
      "grad_norm": 0.39259546995162964,
      "learning_rate": 2e-05,
      "loss": 0.7308,
      "step": 5270
    },
    {
      "epoch": 0.45112781954887216,
      "grad_norm": 0.4098275601863861,
      "learning_rate": 2e-05,
      "loss": 0.7765,
      "step": 5280
    },
    {
      "epoch": 0.45198222829801776,
      "grad_norm": 0.3891877830028534,
      "learning_rate": 2e-05,
      "loss": 0.7376,
      "step": 5290
    },
    {
      "epoch": 0.45283663704716337,
      "grad_norm": 0.41834568977355957,
      "learning_rate": 2e-05,
      "loss": 0.7566,
      "step": 5300
    },
    {
      "epoch": 0.453691045796309,
      "grad_norm": 0.42429131269454956,
      "learning_rate": 2e-05,
      "loss": 0.7556,
      "step": 5310
    },
    {
      "epoch": 0.45454545454545453,
      "grad_norm": 0.41689616441726685,
      "learning_rate": 2e-05,
      "loss": 0.7629,
      "step": 5320
    },
    {
      "epoch": 0.45539986329460014,
      "grad_norm": 0.4062359929084778,
      "learning_rate": 2e-05,
      "loss": 0.7608,
      "step": 5330
    },
    {
      "epoch": 0.45625427204374575,
      "grad_norm": 0.4037449061870575,
      "learning_rate": 2e-05,
      "loss": 0.7616,
      "step": 5340
    },
    {
      "epoch": 0.4571086807928913,
      "grad_norm": 0.3997521698474884,
      "learning_rate": 2e-05,
      "loss": 0.7725,
      "step": 5350
    },
    {
      "epoch": 0.4579630895420369,
      "grad_norm": 0.4115746021270752,
      "learning_rate": 2e-05,
      "loss": 0.7468,
      "step": 5360
    },
    {
      "epoch": 0.4588174982911825,
      "grad_norm": 0.38424789905548096,
      "learning_rate": 2e-05,
      "loss": 0.7519,
      "step": 5370
    },
    {
      "epoch": 0.4596719070403281,
      "grad_norm": 0.3852064907550812,
      "learning_rate": 2e-05,
      "loss": 0.7428,
      "step": 5380
    },
    {
      "epoch": 0.4605263157894737,
      "grad_norm": 0.3820341229438782,
      "learning_rate": 2e-05,
      "loss": 0.7471,
      "step": 5390
    },
    {
      "epoch": 0.4613807245386193,
      "grad_norm": 0.3783576488494873,
      "learning_rate": 2e-05,
      "loss": 0.7516,
      "step": 5400
    },
    {
      "epoch": 0.4622351332877649,
      "grad_norm": 0.3757966160774231,
      "learning_rate": 2e-05,
      "loss": 0.7523,
      "step": 5410
    },
    {
      "epoch": 0.46308954203691044,
      "grad_norm": 0.4041605591773987,
      "learning_rate": 2e-05,
      "loss": 0.7762,
      "step": 5420
    },
    {
      "epoch": 0.46394395078605605,
      "grad_norm": 0.4003051519393921,
      "learning_rate": 2e-05,
      "loss": 0.7088,
      "step": 5430
    },
    {
      "epoch": 0.46479835953520166,
      "grad_norm": 0.4229845106601715,
      "learning_rate": 2e-05,
      "loss": 0.7661,
      "step": 5440
    },
    {
      "epoch": 0.4656527682843472,
      "grad_norm": 0.41391855478286743,
      "learning_rate": 2e-05,
      "loss": 0.7455,
      "step": 5450
    },
    {
      "epoch": 0.4665071770334928,
      "grad_norm": 0.3977081775665283,
      "learning_rate": 2e-05,
      "loss": 0.7703,
      "step": 5460
    },
    {
      "epoch": 0.4673615857826384,
      "grad_norm": 0.4159933924674988,
      "learning_rate": 2e-05,
      "loss": 0.7694,
      "step": 5470
    },
    {
      "epoch": 0.46821599453178403,
      "grad_norm": 0.3817444145679474,
      "learning_rate": 2e-05,
      "loss": 0.7316,
      "step": 5480
    },
    {
      "epoch": 0.4690704032809296,
      "grad_norm": 0.3979772925376892,
      "learning_rate": 2e-05,
      "loss": 0.7606,
      "step": 5490
    },
    {
      "epoch": 0.4699248120300752,
      "grad_norm": 0.40735316276550293,
      "learning_rate": 2e-05,
      "loss": 0.7528,
      "step": 5500
    },
    {
      "epoch": 0.4707792207792208,
      "grad_norm": 0.3842499256134033,
      "learning_rate": 2e-05,
      "loss": 0.7513,
      "step": 5510
    },
    {
      "epoch": 0.47163362952836635,
      "grad_norm": 0.38862600922584534,
      "learning_rate": 2e-05,
      "loss": 0.7658,
      "step": 5520
    },
    {
      "epoch": 0.47248803827751196,
      "grad_norm": 0.37289759516716003,
      "learning_rate": 2e-05,
      "loss": 0.7403,
      "step": 5530
    },
    {
      "epoch": 0.47334244702665756,
      "grad_norm": 0.3781432509422302,
      "learning_rate": 2e-05,
      "loss": 0.7644,
      "step": 5540
    },
    {
      "epoch": 0.4741968557758031,
      "grad_norm": 0.39092928171157837,
      "learning_rate": 2e-05,
      "loss": 0.7546,
      "step": 5550
    },
    {
      "epoch": 0.4750512645249487,
      "grad_norm": 0.3743812143802643,
      "learning_rate": 2e-05,
      "loss": 0.7183,
      "step": 5560
    },
    {
      "epoch": 0.47590567327409433,
      "grad_norm": 0.3968353867530823,
      "learning_rate": 2e-05,
      "loss": 0.7697,
      "step": 5570
    },
    {
      "epoch": 0.47676008202323994,
      "grad_norm": 0.3868055045604706,
      "learning_rate": 2e-05,
      "loss": 0.7162,
      "step": 5580
    },
    {
      "epoch": 0.4776144907723855,
      "grad_norm": 0.3805411159992218,
      "learning_rate": 2e-05,
      "loss": 0.7645,
      "step": 5590
    },
    {
      "epoch": 0.4784688995215311,
      "grad_norm": 0.4218626022338867,
      "learning_rate": 2e-05,
      "loss": 0.7527,
      "step": 5600
    },
    {
      "epoch": 0.4793233082706767,
      "grad_norm": 0.40314140915870667,
      "learning_rate": 2e-05,
      "loss": 0.7495,
      "step": 5610
    },
    {
      "epoch": 0.48017771701982226,
      "grad_norm": 0.3977786898612976,
      "learning_rate": 2e-05,
      "loss": 0.7683,
      "step": 5620
    },
    {
      "epoch": 0.48103212576896787,
      "grad_norm": 0.3766978979110718,
      "learning_rate": 2e-05,
      "loss": 0.7616,
      "step": 5630
    },
    {
      "epoch": 0.4818865345181135,
      "grad_norm": 0.40517958998680115,
      "learning_rate": 2e-05,
      "loss": 0.7781,
      "step": 5640
    },
    {
      "epoch": 0.4827409432672591,
      "grad_norm": 0.39497700333595276,
      "learning_rate": 2e-05,
      "loss": 0.737,
      "step": 5650
    },
    {
      "epoch": 0.48359535201640463,
      "grad_norm": 0.38074126839637756,
      "learning_rate": 2e-05,
      "loss": 0.7436,
      "step": 5660
    },
    {
      "epoch": 0.48444976076555024,
      "grad_norm": 0.3926882743835449,
      "learning_rate": 2e-05,
      "loss": 0.7572,
      "step": 5670
    },
    {
      "epoch": 0.48530416951469585,
      "grad_norm": 0.4020463824272156,
      "learning_rate": 2e-05,
      "loss": 0.7588,
      "step": 5680
    },
    {
      "epoch": 0.4861585782638414,
      "grad_norm": 0.42287537455558777,
      "learning_rate": 2e-05,
      "loss": 0.7541,
      "step": 5690
    },
    {
      "epoch": 0.487012987012987,
      "grad_norm": 0.41800832748413086,
      "learning_rate": 2e-05,
      "loss": 0.7614,
      "step": 5700
    },
    {
      "epoch": 0.4878673957621326,
      "grad_norm": 0.36173734068870544,
      "learning_rate": 2e-05,
      "loss": 0.7245,
      "step": 5710
    },
    {
      "epoch": 0.48872180451127817,
      "grad_norm": 0.40853822231292725,
      "learning_rate": 2e-05,
      "loss": 0.7493,
      "step": 5720
    },
    {
      "epoch": 0.4895762132604238,
      "grad_norm": 0.4014433026313782,
      "learning_rate": 2e-05,
      "loss": 0.721,
      "step": 5730
    },
    {
      "epoch": 0.4904306220095694,
      "grad_norm": 0.40592145919799805,
      "learning_rate": 2e-05,
      "loss": 0.754,
      "step": 5740
    },
    {
      "epoch": 0.491285030758715,
      "grad_norm": 0.37865617871284485,
      "learning_rate": 2e-05,
      "loss": 0.7323,
      "step": 5750
    },
    {
      "epoch": 0.49213943950786054,
      "grad_norm": 0.48795995116233826,
      "learning_rate": 2e-05,
      "loss": 0.7589,
      "step": 5760
    },
    {
      "epoch": 0.49299384825700615,
      "grad_norm": 0.41184067726135254,
      "learning_rate": 2e-05,
      "loss": 0.7679,
      "step": 5770
    },
    {
      "epoch": 0.49384825700615176,
      "grad_norm": 0.43545275926589966,
      "learning_rate": 2e-05,
      "loss": 0.7749,
      "step": 5780
    },
    {
      "epoch": 0.4947026657552973,
      "grad_norm": 0.38941696286201477,
      "learning_rate": 2e-05,
      "loss": 0.7595,
      "step": 5790
    },
    {
      "epoch": 0.4955570745044429,
      "grad_norm": 0.43213599920272827,
      "learning_rate": 2e-05,
      "loss": 0.7534,
      "step": 5800
    },
    {
      "epoch": 0.4964114832535885,
      "grad_norm": 0.3706319034099579,
      "learning_rate": 2e-05,
      "loss": 0.7199,
      "step": 5810
    },
    {
      "epoch": 0.49726589200273413,
      "grad_norm": 0.4167959988117218,
      "learning_rate": 2e-05,
      "loss": 0.7518,
      "step": 5820
    },
    {
      "epoch": 0.4981203007518797,
      "grad_norm": 0.37572261691093445,
      "learning_rate": 2e-05,
      "loss": 0.7241,
      "step": 5830
    },
    {
      "epoch": 0.4989747095010253,
      "grad_norm": 0.39268580079078674,
      "learning_rate": 2e-05,
      "loss": 0.7352,
      "step": 5840
    },
    {
      "epoch": 0.4998291182501709,
      "grad_norm": 0.40648385882377625,
      "learning_rate": 2e-05,
      "loss": 0.7557,
      "step": 5850
    },
    {
      "epoch": 0.5006835269993165,
      "grad_norm": 0.37238332629203796,
      "learning_rate": 2e-05,
      "loss": 0.7373,
      "step": 5860
    },
    {
      "epoch": 0.5015379357484621,
      "grad_norm": 0.3818080425262451,
      "learning_rate": 2e-05,
      "loss": 0.7336,
      "step": 5870
    },
    {
      "epoch": 0.5023923444976076,
      "grad_norm": 0.3863541781902313,
      "learning_rate": 2e-05,
      "loss": 0.7694,
      "step": 5880
    },
    {
      "epoch": 0.5032467532467533,
      "grad_norm": 0.4712207615375519,
      "learning_rate": 2e-05,
      "loss": 0.7497,
      "step": 5890
    },
    {
      "epoch": 0.5041011619958988,
      "grad_norm": 0.4015129506587982,
      "learning_rate": 2e-05,
      "loss": 0.7326,
      "step": 5900
    },
    {
      "epoch": 0.5049555707450444,
      "grad_norm": 0.40788766741752625,
      "learning_rate": 2e-05,
      "loss": 0.7435,
      "step": 5910
    },
    {
      "epoch": 0.50580997949419,
      "grad_norm": 0.3826226592063904,
      "learning_rate": 2e-05,
      "loss": 0.73,
      "step": 5920
    },
    {
      "epoch": 0.5066643882433356,
      "grad_norm": 0.37159624695777893,
      "learning_rate": 2e-05,
      "loss": 0.7545,
      "step": 5930
    },
    {
      "epoch": 0.5075187969924813,
      "grad_norm": 0.39386260509490967,
      "learning_rate": 2e-05,
      "loss": 0.7625,
      "step": 5940
    },
    {
      "epoch": 0.5083732057416268,
      "grad_norm": 0.3973214626312256,
      "learning_rate": 2e-05,
      "loss": 0.7364,
      "step": 5950
    },
    {
      "epoch": 0.5092276144907724,
      "grad_norm": 0.3795078694820404,
      "learning_rate": 2e-05,
      "loss": 0.7633,
      "step": 5960
    },
    {
      "epoch": 0.510082023239918,
      "grad_norm": 0.39751723408699036,
      "learning_rate": 2e-05,
      "loss": 0.7545,
      "step": 5970
    },
    {
      "epoch": 0.5109364319890636,
      "grad_norm": 0.4226236939430237,
      "learning_rate": 2e-05,
      "loss": 0.7425,
      "step": 5980
    },
    {
      "epoch": 0.5117908407382091,
      "grad_norm": 0.3653227686882019,
      "learning_rate": 2e-05,
      "loss": 0.7446,
      "step": 5990
    },
    {
      "epoch": 0.5126452494873548,
      "grad_norm": 0.3984850347042084,
      "learning_rate": 2e-05,
      "loss": 0.7531,
      "step": 6000
    },
    {
      "epoch": 0.5134996582365003,
      "grad_norm": 0.428719162940979,
      "learning_rate": 2e-05,
      "loss": 0.7461,
      "step": 6010
    },
    {
      "epoch": 0.5143540669856459,
      "grad_norm": 0.39256715774536133,
      "learning_rate": 2e-05,
      "loss": 0.7408,
      "step": 6020
    },
    {
      "epoch": 0.5152084757347916,
      "grad_norm": 0.4053582549095154,
      "learning_rate": 2e-05,
      "loss": 0.7576,
      "step": 6030
    },
    {
      "epoch": 0.5160628844839371,
      "grad_norm": 0.39400947093963623,
      "learning_rate": 2e-05,
      "loss": 0.7201,
      "step": 6040
    },
    {
      "epoch": 0.5169172932330827,
      "grad_norm": 0.43286824226379395,
      "learning_rate": 2e-05,
      "loss": 0.7508,
      "step": 6050
    },
    {
      "epoch": 0.5177717019822283,
      "grad_norm": 0.4271588623523712,
      "learning_rate": 2e-05,
      "loss": 0.7406,
      "step": 6060
    },
    {
      "epoch": 0.5186261107313739,
      "grad_norm": 0.4380304217338562,
      "learning_rate": 2e-05,
      "loss": 0.7349,
      "step": 6070
    },
    {
      "epoch": 0.5194805194805194,
      "grad_norm": 0.40119221806526184,
      "learning_rate": 2e-05,
      "loss": 0.7721,
      "step": 6080
    },
    {
      "epoch": 0.5203349282296651,
      "grad_norm": 0.39615076780319214,
      "learning_rate": 2e-05,
      "loss": 0.7289,
      "step": 6090
    },
    {
      "epoch": 0.5211893369788106,
      "grad_norm": 0.39204469323158264,
      "learning_rate": 2e-05,
      "loss": 0.7245,
      "step": 6100
    },
    {
      "epoch": 0.5220437457279562,
      "grad_norm": 0.3824140131473541,
      "learning_rate": 2e-05,
      "loss": 0.7479,
      "step": 6110
    },
    {
      "epoch": 0.5228981544771019,
      "grad_norm": 0.3810947835445404,
      "learning_rate": 2e-05,
      "loss": 0.7244,
      "step": 6120
    },
    {
      "epoch": 0.5237525632262474,
      "grad_norm": 0.39731109142303467,
      "learning_rate": 2e-05,
      "loss": 0.7314,
      "step": 6130
    },
    {
      "epoch": 0.5246069719753931,
      "grad_norm": 0.4059075117111206,
      "learning_rate": 2e-05,
      "loss": 0.7476,
      "step": 6140
    },
    {
      "epoch": 0.5254613807245386,
      "grad_norm": 0.3779563009738922,
      "learning_rate": 2e-05,
      "loss": 0.7467,
      "step": 6150
    },
    {
      "epoch": 0.5263157894736842,
      "grad_norm": 0.4544978737831116,
      "learning_rate": 2e-05,
      "loss": 0.7681,
      "step": 6160
    },
    {
      "epoch": 0.5271701982228298,
      "grad_norm": 0.37293022871017456,
      "learning_rate": 2e-05,
      "loss": 0.7232,
      "step": 6170
    },
    {
      "epoch": 0.5280246069719754,
      "grad_norm": 0.4267317056655884,
      "learning_rate": 2e-05,
      "loss": 0.7331,
      "step": 6180
    },
    {
      "epoch": 0.528879015721121,
      "grad_norm": 0.3786907494068146,
      "learning_rate": 2e-05,
      "loss": 0.7604,
      "step": 6190
    },
    {
      "epoch": 0.5297334244702666,
      "grad_norm": 0.37749990820884705,
      "learning_rate": 2e-05,
      "loss": 0.7344,
      "step": 6200
    },
    {
      "epoch": 0.5305878332194122,
      "grad_norm": 0.39528048038482666,
      "learning_rate": 2e-05,
      "loss": 0.7337,
      "step": 6210
    },
    {
      "epoch": 0.5314422419685577,
      "grad_norm": 0.4105382263660431,
      "learning_rate": 2e-05,
      "loss": 0.7485,
      "step": 6220
    },
    {
      "epoch": 0.5322966507177034,
      "grad_norm": 0.45637255907058716,
      "learning_rate": 2e-05,
      "loss": 0.752,
      "step": 6230
    },
    {
      "epoch": 0.5331510594668489,
      "grad_norm": 0.3753946125507355,
      "learning_rate": 2e-05,
      "loss": 0.7356,
      "step": 6240
    },
    {
      "epoch": 0.5340054682159945,
      "grad_norm": 0.3918614089488983,
      "learning_rate": 2e-05,
      "loss": 0.7297,
      "step": 6250
    },
    {
      "epoch": 0.5348598769651401,
      "grad_norm": 0.4491317570209503,
      "learning_rate": 2e-05,
      "loss": 0.7303,
      "step": 6260
    },
    {
      "epoch": 0.5357142857142857,
      "grad_norm": 0.41136041283607483,
      "learning_rate": 2e-05,
      "loss": 0.7363,
      "step": 6270
    },
    {
      "epoch": 0.5365686944634313,
      "grad_norm": 0.3801671266555786,
      "learning_rate": 2e-05,
      "loss": 0.7583,
      "step": 6280
    },
    {
      "epoch": 0.5374231032125769,
      "grad_norm": 0.36690691113471985,
      "learning_rate": 2e-05,
      "loss": 0.7583,
      "step": 6290
    },
    {
      "epoch": 0.5382775119617225,
      "grad_norm": 0.387388676404953,
      "learning_rate": 2e-05,
      "loss": 0.7498,
      "step": 6300
    },
    {
      "epoch": 0.5391319207108681,
      "grad_norm": 0.39143040776252747,
      "learning_rate": 2e-05,
      "loss": 0.722,
      "step": 6310
    },
    {
      "epoch": 0.5399863294600137,
      "grad_norm": 0.390265554189682,
      "learning_rate": 2e-05,
      "loss": 0.7563,
      "step": 6320
    },
    {
      "epoch": 0.5408407382091592,
      "grad_norm": 0.396502822637558,
      "learning_rate": 2e-05,
      "loss": 0.7447,
      "step": 6330
    },
    {
      "epoch": 0.5416951469583049,
      "grad_norm": 0.3877663016319275,
      "learning_rate": 2e-05,
      "loss": 0.7712,
      "step": 6340
    },
    {
      "epoch": 0.5425495557074504,
      "grad_norm": 0.39390575885772705,
      "learning_rate": 2e-05,
      "loss": 0.7389,
      "step": 6350
    },
    {
      "epoch": 0.543403964456596,
      "grad_norm": 0.4055901765823364,
      "learning_rate": 2e-05,
      "loss": 0.7377,
      "step": 6360
    },
    {
      "epoch": 0.5442583732057417,
      "grad_norm": 0.3841314911842346,
      "learning_rate": 2e-05,
      "loss": 0.71,
      "step": 6370
    },
    {
      "epoch": 0.5451127819548872,
      "grad_norm": 0.418535053730011,
      "learning_rate": 2e-05,
      "loss": 0.7253,
      "step": 6380
    },
    {
      "epoch": 0.5459671907040328,
      "grad_norm": 0.39790672063827515,
      "learning_rate": 2e-05,
      "loss": 0.731,
      "step": 6390
    },
    {
      "epoch": 0.5468215994531784,
      "grad_norm": 0.42406052350997925,
      "learning_rate": 2e-05,
      "loss": 0.7639,
      "step": 6400
    },
    {
      "epoch": 0.547676008202324,
      "grad_norm": 0.3910660147666931,
      "learning_rate": 2e-05,
      "loss": 0.727,
      "step": 6410
    },
    {
      "epoch": 0.5485304169514695,
      "grad_norm": 0.4025813639163971,
      "learning_rate": 2e-05,
      "loss": 0.7592,
      "step": 6420
    },
    {
      "epoch": 0.5493848257006152,
      "grad_norm": 0.5984094738960266,
      "learning_rate": 2e-05,
      "loss": 0.7449,
      "step": 6430
    },
    {
      "epoch": 0.5502392344497608,
      "grad_norm": 0.3710811734199524,
      "learning_rate": 2e-05,
      "loss": 0.7617,
      "step": 6440
    },
    {
      "epoch": 0.5510936431989063,
      "grad_norm": 0.43746209144592285,
      "learning_rate": 2e-05,
      "loss": 0.7719,
      "step": 6450
    },
    {
      "epoch": 0.551948051948052,
      "grad_norm": 0.4054732024669647,
      "learning_rate": 2e-05,
      "loss": 0.7309,
      "step": 6460
    },
    {
      "epoch": 0.5528024606971975,
      "grad_norm": 0.4121415615081787,
      "learning_rate": 2e-05,
      "loss": 0.7456,
      "step": 6470
    },
    {
      "epoch": 0.5536568694463432,
      "grad_norm": 0.4195259213447571,
      "learning_rate": 2e-05,
      "loss": 0.7458,
      "step": 6480
    },
    {
      "epoch": 0.5545112781954887,
      "grad_norm": 0.401277095079422,
      "learning_rate": 2e-05,
      "loss": 0.7502,
      "step": 6490
    },
    {
      "epoch": 0.5553656869446343,
      "grad_norm": 0.39501333236694336,
      "learning_rate": 2e-05,
      "loss": 0.7351,
      "step": 6500
    },
    {
      "epoch": 0.55622009569378,
      "grad_norm": 0.39362087845802307,
      "learning_rate": 2e-05,
      "loss": 0.7487,
      "step": 6510
    },
    {
      "epoch": 0.5570745044429255,
      "grad_norm": 0.3814476430416107,
      "learning_rate": 2e-05,
      "loss": 0.7292,
      "step": 6520
    },
    {
      "epoch": 0.557928913192071,
      "grad_norm": 0.3918703496456146,
      "learning_rate": 2e-05,
      "loss": 0.7161,
      "step": 6530
    },
    {
      "epoch": 0.5587833219412167,
      "grad_norm": 0.39401954412460327,
      "learning_rate": 2e-05,
      "loss": 0.7236,
      "step": 6540
    },
    {
      "epoch": 0.5596377306903623,
      "grad_norm": 0.36988961696624756,
      "learning_rate": 2e-05,
      "loss": 0.7401,
      "step": 6550
    },
    {
      "epoch": 0.5604921394395078,
      "grad_norm": 0.39530149102211,
      "learning_rate": 2e-05,
      "loss": 0.7583,
      "step": 6560
    },
    {
      "epoch": 0.5613465481886535,
      "grad_norm": 0.37524789571762085,
      "learning_rate": 2e-05,
      "loss": 0.7563,
      "step": 6570
    },
    {
      "epoch": 0.562200956937799,
      "grad_norm": 0.38217321038246155,
      "learning_rate": 2e-05,
      "loss": 0.7319,
      "step": 6580
    },
    {
      "epoch": 0.5630553656869446,
      "grad_norm": 0.38432765007019043,
      "learning_rate": 2e-05,
      "loss": 0.768,
      "step": 6590
    },
    {
      "epoch": 0.5639097744360902,
      "grad_norm": 0.40665632486343384,
      "learning_rate": 2e-05,
      "loss": 0.733,
      "step": 6600
    },
    {
      "epoch": 0.5647641831852358,
      "grad_norm": 0.38076648116111755,
      "learning_rate": 2e-05,
      "loss": 0.7437,
      "step": 6610
    },
    {
      "epoch": 0.5656185919343814,
      "grad_norm": 0.411256343126297,
      "learning_rate": 2e-05,
      "loss": 0.7502,
      "step": 6620
    },
    {
      "epoch": 0.566473000683527,
      "grad_norm": 0.3712559938430786,
      "learning_rate": 2e-05,
      "loss": 0.7554,
      "step": 6630
    },
    {
      "epoch": 0.5673274094326726,
      "grad_norm": 0.4126828610897064,
      "learning_rate": 2e-05,
      "loss": 0.7456,
      "step": 6640
    },
    {
      "epoch": 0.5681818181818182,
      "grad_norm": 0.396106094121933,
      "learning_rate": 2e-05,
      "loss": 0.739,
      "step": 6650
    },
    {
      "epoch": 0.5690362269309638,
      "grad_norm": 0.38243263959884644,
      "learning_rate": 2e-05,
      "loss": 0.7322,
      "step": 6660
    },
    {
      "epoch": 0.5698906356801093,
      "grad_norm": 0.38476720452308655,
      "learning_rate": 2e-05,
      "loss": 0.7483,
      "step": 6670
    },
    {
      "epoch": 0.570745044429255,
      "grad_norm": 0.3957832455635071,
      "learning_rate": 2e-05,
      "loss": 0.6942,
      "step": 6680
    },
    {
      "epoch": 0.5715994531784006,
      "grad_norm": 0.37650221586227417,
      "learning_rate": 2e-05,
      "loss": 0.7298,
      "step": 6690
    },
    {
      "epoch": 0.5724538619275461,
      "grad_norm": 0.412524938583374,
      "learning_rate": 2e-05,
      "loss": 0.7503,
      "step": 6700
    },
    {
      "epoch": 0.5733082706766918,
      "grad_norm": 0.40005236864089966,
      "learning_rate": 2e-05,
      "loss": 0.7515,
      "step": 6710
    },
    {
      "epoch": 0.5741626794258373,
      "grad_norm": 0.41363996267318726,
      "learning_rate": 2e-05,
      "loss": 0.736,
      "step": 6720
    },
    {
      "epoch": 0.5750170881749829,
      "grad_norm": 0.39776405692100525,
      "learning_rate": 2e-05,
      "loss": 0.7692,
      "step": 6730
    },
    {
      "epoch": 0.5758714969241285,
      "grad_norm": 0.39384734630584717,
      "learning_rate": 2e-05,
      "loss": 0.749,
      "step": 6740
    },
    {
      "epoch": 0.5767259056732741,
      "grad_norm": 0.4256645143032074,
      "learning_rate": 2e-05,
      "loss": 0.7143,
      "step": 6750
    },
    {
      "epoch": 0.5775803144224196,
      "grad_norm": 0.4060010612010956,
      "learning_rate": 2e-05,
      "loss": 0.738,
      "step": 6760
    },
    {
      "epoch": 0.5784347231715653,
      "grad_norm": 0.39823290705680847,
      "learning_rate": 2e-05,
      "loss": 0.7646,
      "step": 6770
    },
    {
      "epoch": 0.5792891319207109,
      "grad_norm": 0.36494705080986023,
      "learning_rate": 2e-05,
      "loss": 0.7695,
      "step": 6780
    },
    {
      "epoch": 0.5801435406698564,
      "grad_norm": 0.3825265169143677,
      "learning_rate": 2e-05,
      "loss": 0.7362,
      "step": 6790
    },
    {
      "epoch": 0.5809979494190021,
      "grad_norm": 0.3755421042442322,
      "learning_rate": 2e-05,
      "loss": 0.7233,
      "step": 6800
    },
    {
      "epoch": 0.5818523581681476,
      "grad_norm": 0.38475579023361206,
      "learning_rate": 2e-05,
      "loss": 0.7077,
      "step": 6810
    },
    {
      "epoch": 0.5827067669172933,
      "grad_norm": 0.3780248761177063,
      "learning_rate": 2e-05,
      "loss": 0.7491,
      "step": 6820
    },
    {
      "epoch": 0.5835611756664388,
      "grad_norm": 0.4144502282142639,
      "learning_rate": 2e-05,
      "loss": 0.7539,
      "step": 6830
    },
    {
      "epoch": 0.5844155844155844,
      "grad_norm": 0.4001186192035675,
      "learning_rate": 2e-05,
      "loss": 0.7678,
      "step": 6840
    },
    {
      "epoch": 0.58526999316473,
      "grad_norm": 0.4112628102302551,
      "learning_rate": 2e-05,
      "loss": 0.7534,
      "step": 6850
    },
    {
      "epoch": 0.5861244019138756,
      "grad_norm": 0.37572526931762695,
      "learning_rate": 2e-05,
      "loss": 0.75,
      "step": 6860
    },
    {
      "epoch": 0.5869788106630212,
      "grad_norm": 0.37607231736183167,
      "learning_rate": 2e-05,
      "loss": 0.7425,
      "step": 6870
    },
    {
      "epoch": 0.5878332194121668,
      "grad_norm": 0.39626801013946533,
      "learning_rate": 2e-05,
      "loss": 0.7338,
      "step": 6880
    },
    {
      "epoch": 0.5886876281613124,
      "grad_norm": 0.3692607879638672,
      "learning_rate": 2e-05,
      "loss": 0.7155,
      "step": 6890
    },
    {
      "epoch": 0.5895420369104579,
      "grad_norm": 0.40100300312042236,
      "learning_rate": 2e-05,
      "loss": 0.7428,
      "step": 6900
    },
    {
      "epoch": 0.5903964456596036,
      "grad_norm": 0.38678303360939026,
      "learning_rate": 2e-05,
      "loss": 0.743,
      "step": 6910
    },
    {
      "epoch": 0.5912508544087491,
      "grad_norm": 0.39802461862564087,
      "learning_rate": 2e-05,
      "loss": 0.7413,
      "step": 6920
    },
    {
      "epoch": 0.5921052631578947,
      "grad_norm": 0.4174688160419464,
      "learning_rate": 2e-05,
      "loss": 0.7511,
      "step": 6930
    },
    {
      "epoch": 0.5929596719070404,
      "grad_norm": 0.41276833415031433,
      "learning_rate": 2e-05,
      "loss": 0.7311,
      "step": 6940
    },
    {
      "epoch": 0.5938140806561859,
      "grad_norm": 0.4122067093849182,
      "learning_rate": 2e-05,
      "loss": 0.7314,
      "step": 6950
    },
    {
      "epoch": 0.5946684894053315,
      "grad_norm": 0.3720877468585968,
      "learning_rate": 2e-05,
      "loss": 0.7485,
      "step": 6960
    },
    {
      "epoch": 0.5955228981544771,
      "grad_norm": 0.41552406549453735,
      "learning_rate": 2e-05,
      "loss": 0.7434,
      "step": 6970
    },
    {
      "epoch": 0.5963773069036227,
      "grad_norm": 0.3968900740146637,
      "learning_rate": 2e-05,
      "loss": 0.7511,
      "step": 6980
    },
    {
      "epoch": 0.5972317156527683,
      "grad_norm": 0.40619030594825745,
      "learning_rate": 2e-05,
      "loss": 0.7312,
      "step": 6990
    },
    {
      "epoch": 0.5980861244019139,
      "grad_norm": 0.37865859270095825,
      "learning_rate": 2e-05,
      "loss": 0.7393,
      "step": 7000
    },
    {
      "epoch": 0.5989405331510594,
      "grad_norm": 0.4047948122024536,
      "learning_rate": 2e-05,
      "loss": 0.7256,
      "step": 7010
    },
    {
      "epoch": 0.5997949419002051,
      "grad_norm": 0.4093959331512451,
      "learning_rate": 2e-05,
      "loss": 0.7551,
      "step": 7020
    },
    {
      "epoch": 0.6006493506493507,
      "grad_norm": 0.36552152037620544,
      "learning_rate": 2e-05,
      "loss": 0.7174,
      "step": 7030
    },
    {
      "epoch": 0.6015037593984962,
      "grad_norm": 0.39407584071159363,
      "learning_rate": 2e-05,
      "loss": 0.7381,
      "step": 7040
    },
    {
      "epoch": 0.6023581681476419,
      "grad_norm": 0.4003411531448364,
      "learning_rate": 2e-05,
      "loss": 0.7303,
      "step": 7050
    },
    {
      "epoch": 0.6032125768967874,
      "grad_norm": 0.3859708309173584,
      "learning_rate": 2e-05,
      "loss": 0.7668,
      "step": 7060
    },
    {
      "epoch": 0.604066985645933,
      "grad_norm": 0.40104642510414124,
      "learning_rate": 2e-05,
      "loss": 0.757,
      "step": 7070
    },
    {
      "epoch": 0.6049213943950786,
      "grad_norm": 0.3722524046897888,
      "learning_rate": 2e-05,
      "loss": 0.7326,
      "step": 7080
    },
    {
      "epoch": 0.6057758031442242,
      "grad_norm": 0.409563273191452,
      "learning_rate": 2e-05,
      "loss": 0.7526,
      "step": 7090
    },
    {
      "epoch": 0.6066302118933697,
      "grad_norm": 0.38845789432525635,
      "learning_rate": 2e-05,
      "loss": 0.7307,
      "step": 7100
    },
    {
      "epoch": 0.6074846206425154,
      "grad_norm": 0.428221732378006,
      "learning_rate": 2e-05,
      "loss": 0.719,
      "step": 7110
    },
    {
      "epoch": 0.608339029391661,
      "grad_norm": 0.4110715389251709,
      "learning_rate": 2e-05,
      "loss": 0.7534,
      "step": 7120
    },
    {
      "epoch": 0.6091934381408065,
      "grad_norm": 0.4356886148452759,
      "learning_rate": 2e-05,
      "loss": 0.7555,
      "step": 7130
    },
    {
      "epoch": 0.6100478468899522,
      "grad_norm": 0.38887014985084534,
      "learning_rate": 2e-05,
      "loss": 0.7361,
      "step": 7140
    },
    {
      "epoch": 0.6109022556390977,
      "grad_norm": 0.3759215176105499,
      "learning_rate": 2e-05,
      "loss": 0.7481,
      "step": 7150
    },
    {
      "epoch": 0.6117566643882434,
      "grad_norm": 0.38678449392318726,
      "learning_rate": 2e-05,
      "loss": 0.7428,
      "step": 7160
    },
    {
      "epoch": 0.6126110731373889,
      "grad_norm": 0.41070356965065,
      "learning_rate": 2e-05,
      "loss": 0.7305,
      "step": 7170
    },
    {
      "epoch": 0.6134654818865345,
      "grad_norm": 0.3750089108943939,
      "learning_rate": 2e-05,
      "loss": 0.7473,
      "step": 7180
    },
    {
      "epoch": 0.6143198906356802,
      "grad_norm": 0.38430655002593994,
      "learning_rate": 2e-05,
      "loss": 0.7342,
      "step": 7190
    },
    {
      "epoch": 0.6151742993848257,
      "grad_norm": 0.42201271653175354,
      "learning_rate": 2e-05,
      "loss": 0.7519,
      "step": 7200
    },
    {
      "epoch": 0.6160287081339713,
      "grad_norm": 0.3963432312011719,
      "learning_rate": 2e-05,
      "loss": 0.7369,
      "step": 7210
    },
    {
      "epoch": 0.6168831168831169,
      "grad_norm": 0.41454413533210754,
      "learning_rate": 2e-05,
      "loss": 0.7404,
      "step": 7220
    },
    {
      "epoch": 0.6177375256322625,
      "grad_norm": 0.4344048500061035,
      "learning_rate": 2e-05,
      "loss": 0.7467,
      "step": 7230
    },
    {
      "epoch": 0.618591934381408,
      "grad_norm": 0.36278843879699707,
      "learning_rate": 2e-05,
      "loss": 0.759,
      "step": 7240
    },
    {
      "epoch": 0.6194463431305537,
      "grad_norm": 0.3743455708026886,
      "learning_rate": 2e-05,
      "loss": 0.7446,
      "step": 7250
    },
    {
      "epoch": 0.6203007518796992,
      "grad_norm": 1.2426199913024902,
      "learning_rate": 2e-05,
      "loss": 0.7389,
      "step": 7260
    },
    {
      "epoch": 0.6211551606288448,
      "grad_norm": 0.4120846092700958,
      "learning_rate": 2e-05,
      "loss": 0.7543,
      "step": 7270
    },
    {
      "epoch": 0.6220095693779905,
      "grad_norm": 0.4270692467689514,
      "learning_rate": 2e-05,
      "loss": 0.7546,
      "step": 7280
    },
    {
      "epoch": 0.622863978127136,
      "grad_norm": 0.39463505148887634,
      "learning_rate": 2e-05,
      "loss": 0.7261,
      "step": 7290
    },
    {
      "epoch": 0.6237183868762816,
      "grad_norm": 0.386168897151947,
      "learning_rate": 2e-05,
      "loss": 0.7204,
      "step": 7300
    },
    {
      "epoch": 0.6245727956254272,
      "grad_norm": 0.3893314003944397,
      "learning_rate": 2e-05,
      "loss": 0.7221,
      "step": 7310
    },
    {
      "epoch": 0.6254272043745728,
      "grad_norm": 0.3850814700126648,
      "learning_rate": 2e-05,
      "loss": 0.7497,
      "step": 7320
    },
    {
      "epoch": 0.6262816131237184,
      "grad_norm": 0.39102447032928467,
      "learning_rate": 2e-05,
      "loss": 0.7384,
      "step": 7330
    },
    {
      "epoch": 0.627136021872864,
      "grad_norm": 0.39684391021728516,
      "learning_rate": 2e-05,
      "loss": 0.7211,
      "step": 7340
    },
    {
      "epoch": 0.6279904306220095,
      "grad_norm": 0.545001208782196,
      "learning_rate": 2e-05,
      "loss": 0.7269,
      "step": 7350
    },
    {
      "epoch": 0.6288448393711552,
      "grad_norm": 0.40336546301841736,
      "learning_rate": 2e-05,
      "loss": 0.7291,
      "step": 7360
    },
    {
      "epoch": 0.6296992481203008,
      "grad_norm": 0.3780849575996399,
      "learning_rate": 2e-05,
      "loss": 0.755,
      "step": 7370
    },
    {
      "epoch": 0.6305536568694463,
      "grad_norm": 0.4080323874950409,
      "learning_rate": 2e-05,
      "loss": 0.7043,
      "step": 7380
    },
    {
      "epoch": 0.631408065618592,
      "grad_norm": 0.4115821421146393,
      "learning_rate": 2e-05,
      "loss": 0.7249,
      "step": 7390
    },
    {
      "epoch": 0.6322624743677375,
      "grad_norm": 0.4186968207359314,
      "learning_rate": 2e-05,
      "loss": 0.7446,
      "step": 7400
    },
    {
      "epoch": 0.6331168831168831,
      "grad_norm": 0.428628146648407,
      "learning_rate": 2e-05,
      "loss": 0.7318,
      "step": 7410
    },
    {
      "epoch": 0.6339712918660287,
      "grad_norm": 0.37300747632980347,
      "learning_rate": 2e-05,
      "loss": 0.7375,
      "step": 7420
    },
    {
      "epoch": 0.6348257006151743,
      "grad_norm": 0.4241150915622711,
      "learning_rate": 2e-05,
      "loss": 0.7232,
      "step": 7430
    },
    {
      "epoch": 0.6356801093643198,
      "grad_norm": 0.37318888306617737,
      "learning_rate": 2e-05,
      "loss": 0.7339,
      "step": 7440
    },
    {
      "epoch": 0.6365345181134655,
      "grad_norm": 0.3773367404937744,
      "learning_rate": 2e-05,
      "loss": 0.7215,
      "step": 7450
    },
    {
      "epoch": 0.6373889268626111,
      "grad_norm": 0.40832898020744324,
      "learning_rate": 2e-05,
      "loss": 0.7143,
      "step": 7460
    },
    {
      "epoch": 0.6382433356117566,
      "grad_norm": 0.3923477232456207,
      "learning_rate": 2e-05,
      "loss": 0.7211,
      "step": 7470
    },
    {
      "epoch": 0.6390977443609023,
      "grad_norm": 0.38625144958496094,
      "learning_rate": 2e-05,
      "loss": 0.7137,
      "step": 7480
    },
    {
      "epoch": 0.6399521531100478,
      "grad_norm": 0.3956676423549652,
      "learning_rate": 2e-05,
      "loss": 0.7284,
      "step": 7490
    },
    {
      "epoch": 0.6408065618591935,
      "grad_norm": 0.39544951915740967,
      "learning_rate": 2e-05,
      "loss": 0.7764,
      "step": 7500
    },
    {
      "epoch": 0.641660970608339,
      "grad_norm": 0.37415212392807007,
      "learning_rate": 2e-05,
      "loss": 0.7429,
      "step": 7510
    },
    {
      "epoch": 0.6425153793574846,
      "grad_norm": 0.40753981471061707,
      "learning_rate": 2e-05,
      "loss": 0.7696,
      "step": 7520
    },
    {
      "epoch": 0.6433697881066303,
      "grad_norm": 0.4101279079914093,
      "learning_rate": 2e-05,
      "loss": 0.768,
      "step": 7530
    },
    {
      "epoch": 0.6442241968557758,
      "grad_norm": 0.38831228017807007,
      "learning_rate": 2e-05,
      "loss": 0.7563,
      "step": 7540
    },
    {
      "epoch": 0.6450786056049214,
      "grad_norm": 0.39545825123786926,
      "learning_rate": 2e-05,
      "loss": 0.7134,
      "step": 7550
    },
    {
      "epoch": 0.645933014354067,
      "grad_norm": 0.4067087471485138,
      "learning_rate": 2e-05,
      "loss": 0.7475,
      "step": 7560
    },
    {
      "epoch": 0.6467874231032126,
      "grad_norm": 0.40556320548057556,
      "learning_rate": 2e-05,
      "loss": 0.7234,
      "step": 7570
    },
    {
      "epoch": 0.6476418318523581,
      "grad_norm": 0.49691125750541687,
      "learning_rate": 2e-05,
      "loss": 0.7193,
      "step": 7580
    },
    {
      "epoch": 0.6484962406015038,
      "grad_norm": 0.3866407573223114,
      "learning_rate": 2e-05,
      "loss": 0.7533,
      "step": 7590
    },
    {
      "epoch": 0.6493506493506493,
      "grad_norm": 0.3790484368801117,
      "learning_rate": 2e-05,
      "loss": 0.7295,
      "step": 7600
    },
    {
      "epoch": 0.6502050580997949,
      "grad_norm": 0.3936048448085785,
      "learning_rate": 2e-05,
      "loss": 0.7451,
      "step": 7610
    },
    {
      "epoch": 0.6510594668489406,
      "grad_norm": 0.42603182792663574,
      "learning_rate": 2e-05,
      "loss": 0.7479,
      "step": 7620
    },
    {
      "epoch": 0.6519138755980861,
      "grad_norm": 0.39287272095680237,
      "learning_rate": 2e-05,
      "loss": 0.7362,
      "step": 7630
    },
    {
      "epoch": 0.6527682843472317,
      "grad_norm": 0.9408705234527588,
      "learning_rate": 2e-05,
      "loss": 0.7448,
      "step": 7640
    },
    {
      "epoch": 0.6536226930963773,
      "grad_norm": 0.4159223437309265,
      "learning_rate": 2e-05,
      "loss": 0.7182,
      "step": 7650
    },
    {
      "epoch": 0.6544771018455229,
      "grad_norm": 0.39660367369651794,
      "learning_rate": 2e-05,
      "loss": 0.7732,
      "step": 7660
    },
    {
      "epoch": 0.6553315105946685,
      "grad_norm": 0.40769049525260925,
      "learning_rate": 2e-05,
      "loss": 0.706,
      "step": 7670
    },
    {
      "epoch": 0.6561859193438141,
      "grad_norm": 0.4102623760700226,
      "learning_rate": 2e-05,
      "loss": 0.7384,
      "step": 7680
    },
    {
      "epoch": 0.6570403280929596,
      "grad_norm": 0.3598274290561676,
      "learning_rate": 2e-05,
      "loss": 0.717,
      "step": 7690
    },
    {
      "epoch": 0.6578947368421053,
      "grad_norm": 0.37298935651779175,
      "learning_rate": 2e-05,
      "loss": 0.7275,
      "step": 7700
    },
    {
      "epoch": 0.6587491455912509,
      "grad_norm": 0.40319573879241943,
      "learning_rate": 2e-05,
      "loss": 0.7553,
      "step": 7710
    },
    {
      "epoch": 0.6596035543403964,
      "grad_norm": 0.37285202741622925,
      "learning_rate": 2e-05,
      "loss": 0.7066,
      "step": 7720
    },
    {
      "epoch": 0.6604579630895421,
      "grad_norm": 0.4026056230068207,
      "learning_rate": 2e-05,
      "loss": 0.7491,
      "step": 7730
    },
    {
      "epoch": 0.6613123718386876,
      "grad_norm": 0.3749181032180786,
      "learning_rate": 2e-05,
      "loss": 0.7198,
      "step": 7740
    },
    {
      "epoch": 0.6621667805878332,
      "grad_norm": 0.3787565529346466,
      "learning_rate": 2e-05,
      "loss": 0.7491,
      "step": 7750
    },
    {
      "epoch": 0.6630211893369788,
      "grad_norm": 0.41353681683540344,
      "learning_rate": 2e-05,
      "loss": 0.7226,
      "step": 7760
    },
    {
      "epoch": 0.6638755980861244,
      "grad_norm": 0.38578930497169495,
      "learning_rate": 2e-05,
      "loss": 0.7198,
      "step": 7770
    },
    {
      "epoch": 0.66473000683527,
      "grad_norm": 0.38045066595077515,
      "learning_rate": 2e-05,
      "loss": 0.7291,
      "step": 7780
    },
    {
      "epoch": 0.6655844155844156,
      "grad_norm": 0.39423710107803345,
      "learning_rate": 2e-05,
      "loss": 0.716,
      "step": 7790
    },
    {
      "epoch": 0.6664388243335612,
      "grad_norm": 0.39820045232772827,
      "learning_rate": 2e-05,
      "loss": 0.7363,
      "step": 7800
    },
    {
      "epoch": 0.6672932330827067,
      "grad_norm": 0.40215858817100525,
      "learning_rate": 2e-05,
      "loss": 0.7157,
      "step": 7810
    },
    {
      "epoch": 0.6681476418318524,
      "grad_norm": 0.7514727711677551,
      "learning_rate": 2e-05,
      "loss": 0.7332,
      "step": 7820
    },
    {
      "epoch": 0.6690020505809979,
      "grad_norm": 0.41204139590263367,
      "learning_rate": 2e-05,
      "loss": 0.7433,
      "step": 7830
    },
    {
      "epoch": 0.6698564593301436,
      "grad_norm": 0.4157067835330963,
      "learning_rate": 2e-05,
      "loss": 0.7523,
      "step": 7840
    },
    {
      "epoch": 0.6707108680792891,
      "grad_norm": 0.3934507966041565,
      "learning_rate": 2e-05,
      "loss": 0.6979,
      "step": 7850
    },
    {
      "epoch": 0.6715652768284347,
      "grad_norm": 0.474418967962265,
      "learning_rate": 2e-05,
      "loss": 0.7519,
      "step": 7860
    },
    {
      "epoch": 0.6724196855775804,
      "grad_norm": 0.37026700377464294,
      "learning_rate": 2e-05,
      "loss": 0.729,
      "step": 7870
    },
    {
      "epoch": 0.6732740943267259,
      "grad_norm": 0.3985775113105774,
      "learning_rate": 2e-05,
      "loss": 0.7391,
      "step": 7880
    },
    {
      "epoch": 0.6741285030758715,
      "grad_norm": 0.47035539150238037,
      "learning_rate": 2e-05,
      "loss": 0.7173,
      "step": 7890
    },
    {
      "epoch": 0.6749829118250171,
      "grad_norm": 0.37984657287597656,
      "learning_rate": 2e-05,
      "loss": 0.7579,
      "step": 7900
    },
    {
      "epoch": 0.6758373205741627,
      "grad_norm": 0.39628398418426514,
      "learning_rate": 2e-05,
      "loss": 0.7259,
      "step": 7910
    },
    {
      "epoch": 0.6766917293233082,
      "grad_norm": 0.38672786951065063,
      "learning_rate": 2e-05,
      "loss": 0.7395,
      "step": 7920
    },
    {
      "epoch": 0.6775461380724539,
      "grad_norm": 0.38610613346099854,
      "learning_rate": 2e-05,
      "loss": 0.7523,
      "step": 7930
    },
    {
      "epoch": 0.6784005468215994,
      "grad_norm": 0.41302740573883057,
      "learning_rate": 2e-05,
      "loss": 0.7264,
      "step": 7940
    },
    {
      "epoch": 0.679254955570745,
      "grad_norm": 0.3859224021434784,
      "learning_rate": 2e-05,
      "loss": 0.7347,
      "step": 7950
    },
    {
      "epoch": 0.6801093643198907,
      "grad_norm": 0.4091704189777374,
      "learning_rate": 2e-05,
      "loss": 0.7421,
      "step": 7960
    },
    {
      "epoch": 0.6809637730690362,
      "grad_norm": 0.40090590715408325,
      "learning_rate": 2e-05,
      "loss": 0.7208,
      "step": 7970
    },
    {
      "epoch": 0.6818181818181818,
      "grad_norm": 0.3646441400051117,
      "learning_rate": 2e-05,
      "loss": 0.731,
      "step": 7980
    },
    {
      "epoch": 0.6826725905673274,
      "grad_norm": 0.37039488554000854,
      "learning_rate": 2e-05,
      "loss": 0.7234,
      "step": 7990
    },
    {
      "epoch": 0.683526999316473,
      "grad_norm": 0.3955908417701721,
      "learning_rate": 2e-05,
      "loss": 0.7461,
      "step": 8000
    },
    {
      "epoch": 0.6843814080656186,
      "grad_norm": 0.3928298354148865,
      "learning_rate": 2e-05,
      "loss": 0.7189,
      "step": 8010
    },
    {
      "epoch": 0.6852358168147642,
      "grad_norm": 0.37455350160598755,
      "learning_rate": 2e-05,
      "loss": 0.724,
      "step": 8020
    },
    {
      "epoch": 0.6860902255639098,
      "grad_norm": 0.37139347195625305,
      "learning_rate": 2e-05,
      "loss": 0.723,
      "step": 8030
    },
    {
      "epoch": 0.6869446343130554,
      "grad_norm": 0.35550495982170105,
      "learning_rate": 2e-05,
      "loss": 0.7298,
      "step": 8040
    },
    {
      "epoch": 0.687799043062201,
      "grad_norm": 0.37318629026412964,
      "learning_rate": 2e-05,
      "loss": 0.7135,
      "step": 8050
    },
    {
      "epoch": 0.6886534518113465,
      "grad_norm": 0.44238585233688354,
      "learning_rate": 2e-05,
      "loss": 0.7234,
      "step": 8060
    },
    {
      "epoch": 0.6895078605604922,
      "grad_norm": 0.36458757519721985,
      "learning_rate": 2e-05,
      "loss": 0.7327,
      "step": 8070
    },
    {
      "epoch": 0.6903622693096377,
      "grad_norm": 0.3845953941345215,
      "learning_rate": 2e-05,
      "loss": 0.7239,
      "step": 8080
    },
    {
      "epoch": 0.6912166780587833,
      "grad_norm": 0.39425137639045715,
      "learning_rate": 2e-05,
      "loss": 0.74,
      "step": 8090
    },
    {
      "epoch": 0.692071086807929,
      "grad_norm": 0.376758337020874,
      "learning_rate": 2e-05,
      "loss": 0.7652,
      "step": 8100
    },
    {
      "epoch": 0.6929254955570745,
      "grad_norm": 0.40507400035858154,
      "learning_rate": 2e-05,
      "loss": 0.7536,
      "step": 8110
    },
    {
      "epoch": 0.69377990430622,
      "grad_norm": 0.3799040615558624,
      "learning_rate": 2e-05,
      "loss": 0.7022,
      "step": 8120
    },
    {
      "epoch": 0.6946343130553657,
      "grad_norm": 0.3842785656452179,
      "learning_rate": 2e-05,
      "loss": 0.7226,
      "step": 8130
    },
    {
      "epoch": 0.6954887218045113,
      "grad_norm": 0.39032843708992004,
      "learning_rate": 2e-05,
      "loss": 0.7611,
      "step": 8140
    },
    {
      "epoch": 0.6963431305536568,
      "grad_norm": 0.40588632225990295,
      "learning_rate": 2e-05,
      "loss": 0.7561,
      "step": 8150
    },
    {
      "epoch": 0.6971975393028025,
      "grad_norm": 0.453239768743515,
      "learning_rate": 2e-05,
      "loss": 0.6983,
      "step": 8160
    },
    {
      "epoch": 0.698051948051948,
      "grad_norm": 0.3863690197467804,
      "learning_rate": 2e-05,
      "loss": 0.74,
      "step": 8170
    },
    {
      "epoch": 0.6989063568010937,
      "grad_norm": 0.36582303047180176,
      "learning_rate": 2e-05,
      "loss": 0.7022,
      "step": 8180
    },
    {
      "epoch": 0.6997607655502392,
      "grad_norm": 0.40633904933929443,
      "learning_rate": 2e-05,
      "loss": 0.7132,
      "step": 8190
    },
    {
      "epoch": 0.7006151742993848,
      "grad_norm": 0.372663676738739,
      "learning_rate": 2e-05,
      "loss": 0.7304,
      "step": 8200
    },
    {
      "epoch": 0.7014695830485305,
      "grad_norm": 0.36462998390197754,
      "learning_rate": 2e-05,
      "loss": 0.7173,
      "step": 8210
    },
    {
      "epoch": 0.702323991797676,
      "grad_norm": 0.41522669792175293,
      "learning_rate": 2e-05,
      "loss": 0.7338,
      "step": 8220
    },
    {
      "epoch": 0.7031784005468216,
      "grad_norm": 0.41334083676338196,
      "learning_rate": 2e-05,
      "loss": 0.7432,
      "step": 8230
    },
    {
      "epoch": 0.7040328092959672,
      "grad_norm": 0.39568454027175903,
      "learning_rate": 2e-05,
      "loss": 0.7278,
      "step": 8240
    },
    {
      "epoch": 0.7048872180451128,
      "grad_norm": 0.4312908351421356,
      "learning_rate": 2e-05,
      "loss": 0.7263,
      "step": 8250
    },
    {
      "epoch": 0.7057416267942583,
      "grad_norm": 0.37512627243995667,
      "learning_rate": 2e-05,
      "loss": 0.7439,
      "step": 8260
    },
    {
      "epoch": 0.706596035543404,
      "grad_norm": 0.3989044427871704,
      "learning_rate": 2e-05,
      "loss": 0.6999,
      "step": 8270
    },
    {
      "epoch": 0.7074504442925496,
      "grad_norm": 0.4131651520729065,
      "learning_rate": 2e-05,
      "loss": 0.7272,
      "step": 8280
    },
    {
      "epoch": 0.7083048530416951,
      "grad_norm": 0.36454203724861145,
      "learning_rate": 2e-05,
      "loss": 0.723,
      "step": 8290
    },
    {
      "epoch": 0.7091592617908408,
      "grad_norm": 0.39097365736961365,
      "learning_rate": 2e-05,
      "loss": 0.7299,
      "step": 8300
    },
    {
      "epoch": 0.7100136705399863,
      "grad_norm": 0.37854066491127014,
      "learning_rate": 2e-05,
      "loss": 0.7228,
      "step": 8310
    },
    {
      "epoch": 0.7108680792891319,
      "grad_norm": 0.3789846897125244,
      "learning_rate": 2e-05,
      "loss": 0.7311,
      "step": 8320
    },
    {
      "epoch": 0.7117224880382775,
      "grad_norm": 0.3939513564109802,
      "learning_rate": 2e-05,
      "loss": 0.7131,
      "step": 8330
    },
    {
      "epoch": 0.7125768967874231,
      "grad_norm": 0.4239867925643921,
      "learning_rate": 2e-05,
      "loss": 0.7268,
      "step": 8340
    },
    {
      "epoch": 0.7134313055365687,
      "grad_norm": 0.40399208664894104,
      "learning_rate": 2e-05,
      "loss": 0.7603,
      "step": 8350
    },
    {
      "epoch": 0.7142857142857143,
      "grad_norm": 0.402120441198349,
      "learning_rate": 2e-05,
      "loss": 0.735,
      "step": 8360
    },
    {
      "epoch": 0.7151401230348599,
      "grad_norm": 0.35907503962516785,
      "learning_rate": 2e-05,
      "loss": 0.7051,
      "step": 8370
    },
    {
      "epoch": 0.7159945317840055,
      "grad_norm": 0.4730399549007416,
      "learning_rate": 2e-05,
      "loss": 0.7556,
      "step": 8380
    },
    {
      "epoch": 0.7168489405331511,
      "grad_norm": 0.38933417201042175,
      "learning_rate": 2e-05,
      "loss": 0.7169,
      "step": 8390
    },
    {
      "epoch": 0.7177033492822966,
      "grad_norm": 0.3999645709991455,
      "learning_rate": 2e-05,
      "loss": 0.7549,
      "step": 8400
    },
    {
      "epoch": 0.7185577580314423,
      "grad_norm": 0.39319053292274475,
      "learning_rate": 2e-05,
      "loss": 0.7511,
      "step": 8410
    },
    {
      "epoch": 0.7194121667805878,
      "grad_norm": 0.4035261869430542,
      "learning_rate": 2e-05,
      "loss": 0.73,
      "step": 8420
    },
    {
      "epoch": 0.7202665755297334,
      "grad_norm": 0.39820948243141174,
      "learning_rate": 2e-05,
      "loss": 0.739,
      "step": 8430
    },
    {
      "epoch": 0.721120984278879,
      "grad_norm": 0.4172956943511963,
      "learning_rate": 2e-05,
      "loss": 0.7258,
      "step": 8440
    },
    {
      "epoch": 0.7219753930280246,
      "grad_norm": 0.41653865575790405,
      "learning_rate": 2e-05,
      "loss": 0.7234,
      "step": 8450
    },
    {
      "epoch": 0.7228298017771702,
      "grad_norm": 0.39460790157318115,
      "learning_rate": 2e-05,
      "loss": 0.6775,
      "step": 8460
    },
    {
      "epoch": 0.7236842105263158,
      "grad_norm": 0.38110753893852234,
      "learning_rate": 2e-05,
      "loss": 0.7209,
      "step": 8470
    },
    {
      "epoch": 0.7245386192754614,
      "grad_norm": 0.3837377727031708,
      "learning_rate": 2e-05,
      "loss": 0.7426,
      "step": 8480
    },
    {
      "epoch": 0.7253930280246069,
      "grad_norm": 0.38183313608169556,
      "learning_rate": 2e-05,
      "loss": 0.7195,
      "step": 8490
    },
    {
      "epoch": 0.7262474367737526,
      "grad_norm": 0.5087531805038452,
      "learning_rate": 2e-05,
      "loss": 0.7644,
      "step": 8500
    },
    {
      "epoch": 0.7271018455228981,
      "grad_norm": 0.39735472202301025,
      "learning_rate": 2e-05,
      "loss": 0.7289,
      "step": 8510
    },
    {
      "epoch": 0.7279562542720438,
      "grad_norm": 0.40236926078796387,
      "learning_rate": 2e-05,
      "loss": 0.7109,
      "step": 8520
    },
    {
      "epoch": 0.7288106630211894,
      "grad_norm": 0.39852502942085266,
      "learning_rate": 2e-05,
      "loss": 0.7653,
      "step": 8530
    },
    {
      "epoch": 0.7296650717703349,
      "grad_norm": 0.3781871497631073,
      "learning_rate": 2e-05,
      "loss": 0.7668,
      "step": 8540
    },
    {
      "epoch": 0.7305194805194806,
      "grad_norm": 0.3722122311592102,
      "learning_rate": 2e-05,
      "loss": 0.7112,
      "step": 8550
    },
    {
      "epoch": 0.7313738892686261,
      "grad_norm": 0.41003262996673584,
      "learning_rate": 2e-05,
      "loss": 0.6909,
      "step": 8560
    },
    {
      "epoch": 0.7322282980177717,
      "grad_norm": 0.37100768089294434,
      "learning_rate": 2e-05,
      "loss": 0.7394,
      "step": 8570
    },
    {
      "epoch": 0.7330827067669173,
      "grad_norm": 0.37983179092407227,
      "learning_rate": 2e-05,
      "loss": 0.737,
      "step": 8580
    },
    {
      "epoch": 0.7339371155160629,
      "grad_norm": 0.3720548450946808,
      "learning_rate": 2e-05,
      "loss": 0.7359,
      "step": 8590
    },
    {
      "epoch": 0.7347915242652084,
      "grad_norm": 0.39083731174468994,
      "learning_rate": 2e-05,
      "loss": 0.7254,
      "step": 8600
    },
    {
      "epoch": 0.7356459330143541,
      "grad_norm": 0.3761920928955078,
      "learning_rate": 2e-05,
      "loss": 0.7205,
      "step": 8610
    },
    {
      "epoch": 0.7365003417634997,
      "grad_norm": 0.38086843490600586,
      "learning_rate": 2e-05,
      "loss": 0.7218,
      "step": 8620
    },
    {
      "epoch": 0.7373547505126452,
      "grad_norm": 0.4017062187194824,
      "learning_rate": 2e-05,
      "loss": 0.7327,
      "step": 8630
    },
    {
      "epoch": 0.7382091592617909,
      "grad_norm": 0.3945276141166687,
      "learning_rate": 2e-05,
      "loss": 0.734,
      "step": 8640
    },
    {
      "epoch": 0.7390635680109364,
      "grad_norm": 0.3595212697982788,
      "learning_rate": 2e-05,
      "loss": 0.7145,
      "step": 8650
    },
    {
      "epoch": 0.739917976760082,
      "grad_norm": 0.41444915533065796,
      "learning_rate": 2e-05,
      "loss": 0.7516,
      "step": 8660
    },
    {
      "epoch": 0.7407723855092276,
      "grad_norm": 0.4463752210140228,
      "learning_rate": 2e-05,
      "loss": 0.7305,
      "step": 8670
    },
    {
      "epoch": 0.7416267942583732,
      "grad_norm": 0.37402424216270447,
      "learning_rate": 2e-05,
      "loss": 0.7137,
      "step": 8680
    },
    {
      "epoch": 0.7424812030075187,
      "grad_norm": 0.3785907030105591,
      "learning_rate": 2e-05,
      "loss": 0.73,
      "step": 8690
    },
    {
      "epoch": 0.7433356117566644,
      "grad_norm": 0.36398038268089294,
      "learning_rate": 2e-05,
      "loss": 0.7461,
      "step": 8700
    },
    {
      "epoch": 0.74419002050581,
      "grad_norm": 0.3975678086280823,
      "learning_rate": 2e-05,
      "loss": 0.7224,
      "step": 8710
    },
    {
      "epoch": 0.7450444292549556,
      "grad_norm": 0.48168110847473145,
      "learning_rate": 2e-05,
      "loss": 0.7516,
      "step": 8720
    },
    {
      "epoch": 0.7458988380041012,
      "grad_norm": 0.3805778920650482,
      "learning_rate": 2e-05,
      "loss": 0.7162,
      "step": 8730
    },
    {
      "epoch": 0.7467532467532467,
      "grad_norm": 0.40565064549446106,
      "learning_rate": 2e-05,
      "loss": 0.7347,
      "step": 8740
    },
    {
      "epoch": 0.7476076555023924,
      "grad_norm": 0.40834522247314453,
      "learning_rate": 2e-05,
      "loss": 0.7202,
      "step": 8750
    },
    {
      "epoch": 0.7484620642515379,
      "grad_norm": 0.3792518079280853,
      "learning_rate": 2e-05,
      "loss": 0.7166,
      "step": 8760
    },
    {
      "epoch": 0.7493164730006835,
      "grad_norm": 0.35768914222717285,
      "learning_rate": 2e-05,
      "loss": 0.7361,
      "step": 8770
    },
    {
      "epoch": 0.7501708817498292,
      "grad_norm": 0.38420337438583374,
      "learning_rate": 2e-05,
      "loss": 0.7369,
      "step": 8780
    },
    {
      "epoch": 0.7510252904989747,
      "grad_norm": 0.38742583990097046,
      "learning_rate": 2e-05,
      "loss": 0.741,
      "step": 8790
    },
    {
      "epoch": 0.7518796992481203,
      "grad_norm": 0.4280838072299957,
      "learning_rate": 2e-05,
      "loss": 0.7258,
      "step": 8800
    },
    {
      "epoch": 0.7527341079972659,
      "grad_norm": 0.3843540549278259,
      "learning_rate": 2e-05,
      "loss": 0.7248,
      "step": 8810
    },
    {
      "epoch": 0.7535885167464115,
      "grad_norm": 0.3759978711605072,
      "learning_rate": 2e-05,
      "loss": 0.714,
      "step": 8820
    },
    {
      "epoch": 0.754442925495557,
      "grad_norm": 0.42796790599823,
      "learning_rate": 2e-05,
      "loss": 0.7565,
      "step": 8830
    },
    {
      "epoch": 0.7552973342447027,
      "grad_norm": 0.4456244707107544,
      "learning_rate": 2e-05,
      "loss": 0.752,
      "step": 8840
    },
    {
      "epoch": 0.7561517429938482,
      "grad_norm": 0.37640926241874695,
      "learning_rate": 2e-05,
      "loss": 0.7226,
      "step": 8850
    },
    {
      "epoch": 0.7570061517429938,
      "grad_norm": 0.37429079413414,
      "learning_rate": 2e-05,
      "loss": 0.7139,
      "step": 8860
    },
    {
      "epoch": 0.7578605604921395,
      "grad_norm": 0.36706632375717163,
      "learning_rate": 2e-05,
      "loss": 0.7214,
      "step": 8870
    },
    {
      "epoch": 0.758714969241285,
      "grad_norm": 0.40641772747039795,
      "learning_rate": 2e-05,
      "loss": 0.7242,
      "step": 8880
    },
    {
      "epoch": 0.7595693779904307,
      "grad_norm": 0.3736916482448578,
      "learning_rate": 2e-05,
      "loss": 0.7117,
      "step": 8890
    },
    {
      "epoch": 0.7604237867395762,
      "grad_norm": 0.38514000177383423,
      "learning_rate": 2e-05,
      "loss": 0.7412,
      "step": 8900
    },
    {
      "epoch": 0.7612781954887218,
      "grad_norm": 0.376800000667572,
      "learning_rate": 2e-05,
      "loss": 0.7228,
      "step": 8910
    },
    {
      "epoch": 0.7621326042378674,
      "grad_norm": 0.41300874948501587,
      "learning_rate": 2e-05,
      "loss": 0.7173,
      "step": 8920
    },
    {
      "epoch": 0.762987012987013,
      "grad_norm": 0.37376734614372253,
      "learning_rate": 2e-05,
      "loss": 0.7142,
      "step": 8930
    },
    {
      "epoch": 0.7638414217361585,
      "grad_norm": 0.36455538868904114,
      "learning_rate": 2e-05,
      "loss": 0.7264,
      "step": 8940
    },
    {
      "epoch": 0.7646958304853042,
      "grad_norm": 0.3930019438266754,
      "learning_rate": 2e-05,
      "loss": 0.7646,
      "step": 8950
    },
    {
      "epoch": 0.7655502392344498,
      "grad_norm": 0.9772857427597046,
      "learning_rate": 2e-05,
      "loss": 0.7384,
      "step": 8960
    },
    {
      "epoch": 0.7664046479835953,
      "grad_norm": 0.35807716846466064,
      "learning_rate": 2e-05,
      "loss": 0.7316,
      "step": 8970
    },
    {
      "epoch": 0.767259056732741,
      "grad_norm": 0.3853870630264282,
      "learning_rate": 2e-05,
      "loss": 0.7085,
      "step": 8980
    },
    {
      "epoch": 0.7681134654818865,
      "grad_norm": 0.3780663013458252,
      "learning_rate": 2e-05,
      "loss": 0.7134,
      "step": 8990
    },
    {
      "epoch": 0.7689678742310321,
      "grad_norm": 0.37526047229766846,
      "learning_rate": 2e-05,
      "loss": 0.7158,
      "step": 9000
    },
    {
      "epoch": 0.7698222829801777,
      "grad_norm": 0.409743070602417,
      "learning_rate": 2e-05,
      "loss": 0.7587,
      "step": 9010
    },
    {
      "epoch": 0.7706766917293233,
      "grad_norm": 0.4030572772026062,
      "learning_rate": 2e-05,
      "loss": 0.7276,
      "step": 9020
    },
    {
      "epoch": 0.7715311004784688,
      "grad_norm": 0.39499789476394653,
      "learning_rate": 2e-05,
      "loss": 0.7922,
      "step": 9030
    },
    {
      "epoch": 0.7723855092276145,
      "grad_norm": 0.39042243361473083,
      "learning_rate": 2e-05,
      "loss": 0.7216,
      "step": 9040
    },
    {
      "epoch": 0.7732399179767601,
      "grad_norm": 0.3861633241176605,
      "learning_rate": 2e-05,
      "loss": 0.7016,
      "step": 9050
    },
    {
      "epoch": 0.7740943267259057,
      "grad_norm": 0.4414706528186798,
      "learning_rate": 2e-05,
      "loss": 0.7297,
      "step": 9060
    },
    {
      "epoch": 0.7749487354750513,
      "grad_norm": 0.39177972078323364,
      "learning_rate": 2e-05,
      "loss": 0.7091,
      "step": 9070
    },
    {
      "epoch": 0.7758031442241968,
      "grad_norm": 0.37594228982925415,
      "learning_rate": 2e-05,
      "loss": 0.7188,
      "step": 9080
    },
    {
      "epoch": 0.7766575529733425,
      "grad_norm": 0.39178118109703064,
      "learning_rate": 2e-05,
      "loss": 0.7469,
      "step": 9090
    },
    {
      "epoch": 0.777511961722488,
      "grad_norm": 0.381268709897995,
      "learning_rate": 2e-05,
      "loss": 0.7104,
      "step": 9100
    },
    {
      "epoch": 0.7783663704716336,
      "grad_norm": 0.38252103328704834,
      "learning_rate": 2e-05,
      "loss": 0.7408,
      "step": 9110
    },
    {
      "epoch": 0.7792207792207793,
      "grad_norm": 0.3806038200855255,
      "learning_rate": 2e-05,
      "loss": 0.7389,
      "step": 9120
    },
    {
      "epoch": 0.7800751879699248,
      "grad_norm": 0.40963301062583923,
      "learning_rate": 2e-05,
      "loss": 0.7569,
      "step": 9130
    },
    {
      "epoch": 0.7809295967190704,
      "grad_norm": 0.39828911423683167,
      "learning_rate": 2e-05,
      "loss": 0.7184,
      "step": 9140
    },
    {
      "epoch": 0.781784005468216,
      "grad_norm": 0.3953932523727417,
      "learning_rate": 2e-05,
      "loss": 0.7204,
      "step": 9150
    },
    {
      "epoch": 0.7826384142173616,
      "grad_norm": 0.3855108618736267,
      "learning_rate": 2e-05,
      "loss": 0.7212,
      "step": 9160
    },
    {
      "epoch": 0.7834928229665071,
      "grad_norm": 0.40698444843292236,
      "learning_rate": 2e-05,
      "loss": 0.7221,
      "step": 9170
    },
    {
      "epoch": 0.7843472317156528,
      "grad_norm": 0.4494570195674896,
      "learning_rate": 2e-05,
      "loss": 0.7224,
      "step": 9180
    },
    {
      "epoch": 0.7852016404647983,
      "grad_norm": 0.38095125555992126,
      "learning_rate": 2e-05,
      "loss": 0.7115,
      "step": 9190
    },
    {
      "epoch": 0.7860560492139439,
      "grad_norm": 0.38900408148765564,
      "learning_rate": 2e-05,
      "loss": 0.7521,
      "step": 9200
    },
    {
      "epoch": 0.7869104579630896,
      "grad_norm": 0.38929495215415955,
      "learning_rate": 2e-05,
      "loss": 0.7121,
      "step": 9210
    },
    {
      "epoch": 0.7877648667122351,
      "grad_norm": 0.42904821038246155,
      "learning_rate": 2e-05,
      "loss": 0.7319,
      "step": 9220
    },
    {
      "epoch": 0.7886192754613808,
      "grad_norm": 0.41673603653907776,
      "learning_rate": 2e-05,
      "loss": 0.7428,
      "step": 9230
    },
    {
      "epoch": 0.7894736842105263,
      "grad_norm": 0.377351313829422,
      "learning_rate": 2e-05,
      "loss": 0.7185,
      "step": 9240
    },
    {
      "epoch": 0.7903280929596719,
      "grad_norm": 0.4008667767047882,
      "learning_rate": 2e-05,
      "loss": 0.7304,
      "step": 9250
    },
    {
      "epoch": 0.7911825017088175,
      "grad_norm": 0.40665316581726074,
      "learning_rate": 2e-05,
      "loss": 0.7233,
      "step": 9260
    },
    {
      "epoch": 0.7920369104579631,
      "grad_norm": 0.39836522936820984,
      "learning_rate": 2e-05,
      "loss": 0.692,
      "step": 9270
    },
    {
      "epoch": 0.7928913192071086,
      "grad_norm": 0.4009968936443329,
      "learning_rate": 2e-05,
      "loss": 0.6984,
      "step": 9280
    },
    {
      "epoch": 0.7937457279562543,
      "grad_norm": 0.3839957118034363,
      "learning_rate": 2e-05,
      "loss": 0.7434,
      "step": 9290
    },
    {
      "epoch": 0.7946001367053999,
      "grad_norm": 0.3795250952243805,
      "learning_rate": 2e-05,
      "loss": 0.7265,
      "step": 9300
    },
    {
      "epoch": 0.7954545454545454,
      "grad_norm": 0.4379538893699646,
      "learning_rate": 2e-05,
      "loss": 0.7498,
      "step": 9310
    },
    {
      "epoch": 0.7963089542036911,
      "grad_norm": 0.3773500621318817,
      "learning_rate": 2e-05,
      "loss": 0.7386,
      "step": 9320
    },
    {
      "epoch": 0.7971633629528366,
      "grad_norm": 0.3996424376964569,
      "learning_rate": 2e-05,
      "loss": 0.7413,
      "step": 9330
    },
    {
      "epoch": 0.7980177717019822,
      "grad_norm": 0.38832151889801025,
      "learning_rate": 2e-05,
      "loss": 0.706,
      "step": 9340
    },
    {
      "epoch": 0.7988721804511278,
      "grad_norm": 0.41009020805358887,
      "learning_rate": 2e-05,
      "loss": 0.7428,
      "step": 9350
    },
    {
      "epoch": 0.7997265892002734,
      "grad_norm": 0.42310231924057007,
      "learning_rate": 2e-05,
      "loss": 0.6903,
      "step": 9360
    },
    {
      "epoch": 0.800580997949419,
      "grad_norm": 0.40602195262908936,
      "learning_rate": 2e-05,
      "loss": 0.6964,
      "step": 9370
    },
    {
      "epoch": 0.8014354066985646,
      "grad_norm": 0.35780948400497437,
      "learning_rate": 2e-05,
      "loss": 0.7155,
      "step": 9380
    },
    {
      "epoch": 0.8022898154477102,
      "grad_norm": 0.4097701907157898,
      "learning_rate": 2e-05,
      "loss": 0.7285,
      "step": 9390
    },
    {
      "epoch": 0.8031442241968558,
      "grad_norm": 0.3953598439693451,
      "learning_rate": 2e-05,
      "loss": 0.7211,
      "step": 9400
    },
    {
      "epoch": 0.8039986329460014,
      "grad_norm": 0.41654783487319946,
      "learning_rate": 2e-05,
      "loss": 0.7323,
      "step": 9410
    },
    {
      "epoch": 0.8048530416951469,
      "grad_norm": 0.38317379355430603,
      "learning_rate": 2e-05,
      "loss": 0.7165,
      "step": 9420
    },
    {
      "epoch": 0.8057074504442926,
      "grad_norm": 0.3758383095264435,
      "learning_rate": 2e-05,
      "loss": 0.7391,
      "step": 9430
    },
    {
      "epoch": 0.8065618591934381,
      "grad_norm": 0.37433817982673645,
      "learning_rate": 2e-05,
      "loss": 0.7268,
      "step": 9440
    },
    {
      "epoch": 0.8074162679425837,
      "grad_norm": 0.42244112491607666,
      "learning_rate": 2e-05,
      "loss": 0.7296,
      "step": 9450
    },
    {
      "epoch": 0.8082706766917294,
      "grad_norm": 0.382637619972229,
      "learning_rate": 2e-05,
      "loss": 0.7094,
      "step": 9460
    },
    {
      "epoch": 0.8091250854408749,
      "grad_norm": 0.38067856431007385,
      "learning_rate": 2e-05,
      "loss": 0.6988,
      "step": 9470
    },
    {
      "epoch": 0.8099794941900205,
      "grad_norm": 0.4207744598388672,
      "learning_rate": 2e-05,
      "loss": 0.7278,
      "step": 9480
    },
    {
      "epoch": 0.8108339029391661,
      "grad_norm": 0.46020832657814026,
      "learning_rate": 2e-05,
      "loss": 0.7226,
      "step": 9490
    },
    {
      "epoch": 0.8116883116883117,
      "grad_norm": 0.40805020928382874,
      "learning_rate": 2e-05,
      "loss": 0.7509,
      "step": 9500
    },
    {
      "epoch": 0.8125427204374572,
      "grad_norm": 0.4145103693008423,
      "learning_rate": 2e-05,
      "loss": 0.7392,
      "step": 9510
    },
    {
      "epoch": 0.8133971291866029,
      "grad_norm": 0.4663565158843994,
      "learning_rate": 2e-05,
      "loss": 0.759,
      "step": 9520
    },
    {
      "epoch": 0.8142515379357484,
      "grad_norm": 0.39531323313713074,
      "learning_rate": 2e-05,
      "loss": 0.7389,
      "step": 9530
    },
    {
      "epoch": 0.815105946684894,
      "grad_norm": 0.351619690656662,
      "learning_rate": 2e-05,
      "loss": 0.7104,
      "step": 9540
    },
    {
      "epoch": 0.8159603554340397,
      "grad_norm": 0.40503090620040894,
      "learning_rate": 2e-05,
      "loss": 0.7533,
      "step": 9550
    },
    {
      "epoch": 0.8168147641831852,
      "grad_norm": 0.38226318359375,
      "learning_rate": 2e-05,
      "loss": 0.7357,
      "step": 9560
    },
    {
      "epoch": 0.8176691729323309,
      "grad_norm": 0.3873129189014435,
      "learning_rate": 2e-05,
      "loss": 0.7425,
      "step": 9570
    },
    {
      "epoch": 0.8185235816814764,
      "grad_norm": 0.3682829737663269,
      "learning_rate": 2e-05,
      "loss": 0.7324,
      "step": 9580
    },
    {
      "epoch": 0.819377990430622,
      "grad_norm": 0.39638662338256836,
      "learning_rate": 2e-05,
      "loss": 0.7108,
      "step": 9590
    },
    {
      "epoch": 0.8202323991797676,
      "grad_norm": 0.3786948621273041,
      "learning_rate": 2e-05,
      "loss": 0.7384,
      "step": 9600
    },
    {
      "epoch": 0.8210868079289132,
      "grad_norm": 0.3789300322532654,
      "learning_rate": 2e-05,
      "loss": 0.7137,
      "step": 9610
    },
    {
      "epoch": 0.8219412166780588,
      "grad_norm": 0.3878823518753052,
      "learning_rate": 2e-05,
      "loss": 0.7331,
      "step": 9620
    },
    {
      "epoch": 0.8227956254272044,
      "grad_norm": 0.41125842928886414,
      "learning_rate": 2e-05,
      "loss": 0.7225,
      "step": 9630
    },
    {
      "epoch": 0.82365003417635,
      "grad_norm": 0.3630281388759613,
      "learning_rate": 2e-05,
      "loss": 0.7056,
      "step": 9640
    },
    {
      "epoch": 0.8245044429254955,
      "grad_norm": 0.39767882227897644,
      "learning_rate": 2e-05,
      "loss": 0.7249,
      "step": 9650
    },
    {
      "epoch": 0.8253588516746412,
      "grad_norm": 0.3783910274505615,
      "learning_rate": 2e-05,
      "loss": 0.7202,
      "step": 9660
    },
    {
      "epoch": 0.8262132604237867,
      "grad_norm": 0.3670635521411896,
      "learning_rate": 2e-05,
      "loss": 0.7096,
      "step": 9670
    },
    {
      "epoch": 0.8270676691729323,
      "grad_norm": 0.39036595821380615,
      "learning_rate": 2e-05,
      "loss": 0.739,
      "step": 9680
    },
    {
      "epoch": 0.827922077922078,
      "grad_norm": 0.40312501788139343,
      "learning_rate": 2e-05,
      "loss": 0.7378,
      "step": 9690
    },
    {
      "epoch": 0.8287764866712235,
      "grad_norm": 0.3603912889957428,
      "learning_rate": 2e-05,
      "loss": 0.7168,
      "step": 9700
    },
    {
      "epoch": 0.829630895420369,
      "grad_norm": 0.3715411424636841,
      "learning_rate": 2e-05,
      "loss": 0.725,
      "step": 9710
    },
    {
      "epoch": 0.8304853041695147,
      "grad_norm": 0.37530001997947693,
      "learning_rate": 2e-05,
      "loss": 0.7323,
      "step": 9720
    },
    {
      "epoch": 0.8313397129186603,
      "grad_norm": 0.3843953013420105,
      "learning_rate": 2e-05,
      "loss": 0.7501,
      "step": 9730
    },
    {
      "epoch": 0.8321941216678059,
      "grad_norm": 0.3791062831878662,
      "learning_rate": 2e-05,
      "loss": 0.7333,
      "step": 9740
    },
    {
      "epoch": 0.8330485304169515,
      "grad_norm": 0.3918222188949585,
      "learning_rate": 2e-05,
      "loss": 0.718,
      "step": 9750
    },
    {
      "epoch": 0.833902939166097,
      "grad_norm": 0.39870718121528625,
      "learning_rate": 2e-05,
      "loss": 0.7129,
      "step": 9760
    },
    {
      "epoch": 0.8347573479152427,
      "grad_norm": 0.41660210490226746,
      "learning_rate": 2e-05,
      "loss": 0.75,
      "step": 9770
    },
    {
      "epoch": 0.8356117566643882,
      "grad_norm": 0.39963558316230774,
      "learning_rate": 2e-05,
      "loss": 0.7075,
      "step": 9780
    },
    {
      "epoch": 0.8364661654135338,
      "grad_norm": 0.4176487922668457,
      "learning_rate": 2e-05,
      "loss": 0.695,
      "step": 9790
    },
    {
      "epoch": 0.8373205741626795,
      "grad_norm": 0.3736605942249298,
      "learning_rate": 2e-05,
      "loss": 0.6901,
      "step": 9800
    },
    {
      "epoch": 0.838174982911825,
      "grad_norm": 0.386368066072464,
      "learning_rate": 2e-05,
      "loss": 0.7196,
      "step": 9810
    },
    {
      "epoch": 0.8390293916609706,
      "grad_norm": 0.3772842288017273,
      "learning_rate": 2e-05,
      "loss": 0.7134,
      "step": 9820
    },
    {
      "epoch": 0.8398838004101162,
      "grad_norm": 0.3823299705982208,
      "learning_rate": 2e-05,
      "loss": 0.7005,
      "step": 9830
    },
    {
      "epoch": 0.8407382091592618,
      "grad_norm": 0.38625821471214294,
      "learning_rate": 2e-05,
      "loss": 0.7064,
      "step": 9840
    },
    {
      "epoch": 0.8415926179084073,
      "grad_norm": 0.36672118306159973,
      "learning_rate": 2e-05,
      "loss": 0.7155,
      "step": 9850
    },
    {
      "epoch": 0.842447026657553,
      "grad_norm": 0.40560615062713623,
      "learning_rate": 2e-05,
      "loss": 0.7416,
      "step": 9860
    },
    {
      "epoch": 0.8433014354066986,
      "grad_norm": 0.38724344968795776,
      "learning_rate": 2e-05,
      "loss": 0.7311,
      "step": 9870
    },
    {
      "epoch": 0.8441558441558441,
      "grad_norm": 0.3788376450538635,
      "learning_rate": 2e-05,
      "loss": 0.7079,
      "step": 9880
    },
    {
      "epoch": 0.8450102529049898,
      "grad_norm": 0.3901161551475525,
      "learning_rate": 2e-05,
      "loss": 0.7153,
      "step": 9890
    },
    {
      "epoch": 0.8458646616541353,
      "grad_norm": 0.39159348607063293,
      "learning_rate": 2e-05,
      "loss": 0.7237,
      "step": 9900
    },
    {
      "epoch": 0.846719070403281,
      "grad_norm": 0.34832075238227844,
      "learning_rate": 2e-05,
      "loss": 0.7185,
      "step": 9910
    },
    {
      "epoch": 0.8475734791524265,
      "grad_norm": 0.39487484097480774,
      "learning_rate": 2e-05,
      "loss": 0.7192,
      "step": 9920
    },
    {
      "epoch": 0.8484278879015721,
      "grad_norm": 0.3903859853744507,
      "learning_rate": 2e-05,
      "loss": 0.7068,
      "step": 9930
    },
    {
      "epoch": 0.8492822966507177,
      "grad_norm": 0.37648048996925354,
      "learning_rate": 2e-05,
      "loss": 0.727,
      "step": 9940
    },
    {
      "epoch": 0.8501367053998633,
      "grad_norm": 0.3600155711174011,
      "learning_rate": 2e-05,
      "loss": 0.6932,
      "step": 9950
    },
    {
      "epoch": 0.8509911141490089,
      "grad_norm": 0.4149894714355469,
      "learning_rate": 2e-05,
      "loss": 0.7299,
      "step": 9960
    },
    {
      "epoch": 0.8518455228981545,
      "grad_norm": 0.3715495467185974,
      "learning_rate": 2e-05,
      "loss": 0.7214,
      "step": 9970
    },
    {
      "epoch": 0.8526999316473001,
      "grad_norm": 0.37358298897743225,
      "learning_rate": 2e-05,
      "loss": 0.7032,
      "step": 9980
    },
    {
      "epoch": 0.8535543403964456,
      "grad_norm": 0.3706444203853607,
      "learning_rate": 2e-05,
      "loss": 0.7007,
      "step": 9990
    },
    {
      "epoch": 0.8544087491455913,
      "grad_norm": 0.3661361336708069,
      "learning_rate": 2e-05,
      "loss": 0.7283,
      "step": 10000
    },
    {
      "epoch": 0.8552631578947368,
      "grad_norm": 0.3790668249130249,
      "learning_rate": 2e-05,
      "loss": 0.7009,
      "step": 10010
    },
    {
      "epoch": 0.8561175666438824,
      "grad_norm": 0.3988180458545685,
      "learning_rate": 2e-05,
      "loss": 0.75,
      "step": 10020
    },
    {
      "epoch": 0.856971975393028,
      "grad_norm": 0.37371933460235596,
      "learning_rate": 2e-05,
      "loss": 0.7243,
      "step": 10030
    },
    {
      "epoch": 0.8578263841421736,
      "grad_norm": 0.37391796708106995,
      "learning_rate": 2e-05,
      "loss": 0.7149,
      "step": 10040
    },
    {
      "epoch": 0.8586807928913192,
      "grad_norm": 0.38872241973876953,
      "learning_rate": 2e-05,
      "loss": 0.7207,
      "step": 10050
    },
    {
      "epoch": 0.8595352016404648,
      "grad_norm": 0.36937227845191956,
      "learning_rate": 2e-05,
      "loss": 0.6988,
      "step": 10060
    },
    {
      "epoch": 0.8603896103896104,
      "grad_norm": 0.4042976200580597,
      "learning_rate": 2e-05,
      "loss": 0.7102,
      "step": 10070
    },
    {
      "epoch": 0.861244019138756,
      "grad_norm": 0.3888101577758789,
      "learning_rate": 2e-05,
      "loss": 0.7725,
      "step": 10080
    },
    {
      "epoch": 0.8620984278879016,
      "grad_norm": 1.0467019081115723,
      "learning_rate": 2e-05,
      "loss": 0.7329,
      "step": 10090
    },
    {
      "epoch": 0.8629528366370471,
      "grad_norm": 0.3860304057598114,
      "learning_rate": 2e-05,
      "loss": 0.713,
      "step": 10100
    },
    {
      "epoch": 0.8638072453861928,
      "grad_norm": 0.4702669382095337,
      "learning_rate": 2e-05,
      "loss": 0.7075,
      "step": 10110
    },
    {
      "epoch": 0.8646616541353384,
      "grad_norm": 0.374645471572876,
      "learning_rate": 2e-05,
      "loss": 0.7059,
      "step": 10120
    },
    {
      "epoch": 0.8655160628844839,
      "grad_norm": 0.392675518989563,
      "learning_rate": 2e-05,
      "loss": 0.7246,
      "step": 10130
    },
    {
      "epoch": 0.8663704716336296,
      "grad_norm": 0.44317933917045593,
      "learning_rate": 2e-05,
      "loss": 0.7271,
      "step": 10140
    },
    {
      "epoch": 0.8672248803827751,
      "grad_norm": 0.41116687655448914,
      "learning_rate": 2e-05,
      "loss": 0.7124,
      "step": 10150
    },
    {
      "epoch": 0.8680792891319207,
      "grad_norm": 0.37885040044784546,
      "learning_rate": 2e-05,
      "loss": 0.7022,
      "step": 10160
    },
    {
      "epoch": 0.8689336978810663,
      "grad_norm": 0.3641525208950043,
      "learning_rate": 2e-05,
      "loss": 0.718,
      "step": 10170
    },
    {
      "epoch": 0.8697881066302119,
      "grad_norm": 0.4093606173992157,
      "learning_rate": 1.999957358086795e-05,
      "loss": 0.7389,
      "step": 10180
    },
    {
      "epoch": 0.8706425153793574,
      "grad_norm": 0.38578811287879944,
      "learning_rate": 1.9991993813020483e-05,
      "loss": 0.7136,
      "step": 10190
    },
    {
      "epoch": 0.8714969241285031,
      "grad_norm": 0.39270082116127014,
      "learning_rate": 1.9974946373028628e-05,
      "loss": 0.7495,
      "step": 10200
    },
    {
      "epoch": 0.8723513328776487,
      "grad_norm": 0.38821524381637573,
      "learning_rate": 1.994844749501764e-05,
      "loss": 0.714,
      "step": 10210
    },
    {
      "epoch": 0.8732057416267942,
      "grad_norm": 0.3797816336154938,
      "learning_rate": 1.9912522413632988e-05,
      "loss": 0.7324,
      "step": 10220
    },
    {
      "epoch": 0.8740601503759399,
      "grad_norm": 0.398258239030838,
      "learning_rate": 1.9867205340009656e-05,
      "loss": 0.7259,
      "step": 10230
    },
    {
      "epoch": 0.8749145591250854,
      "grad_norm": 0.3801535964012146,
      "learning_rate": 1.9812539429193164e-05,
      "loss": 0.7052,
      "step": 10240
    },
    {
      "epoch": 0.8757689678742311,
      "grad_norm": 0.3660241663455963,
      "learning_rate": 1.9748576739043423e-05,
      "loss": 0.7133,
      "step": 10250
    },
    {
      "epoch": 0.8766233766233766,
      "grad_norm": 0.42410120368003845,
      "learning_rate": 1.967537818066047e-05,
      "loss": 0.7402,
      "step": 10260
    },
    {
      "epoch": 0.8774777853725222,
      "grad_norm": 0.39106079936027527,
      "learning_rate": 1.959301346037941e-05,
      "loss": 0.7251,
      "step": 10270
    },
    {
      "epoch": 0.8783321941216679,
      "grad_norm": 0.37869200110435486,
      "learning_rate": 1.9501561013389646e-05,
      "loss": 0.7412,
      "step": 10280
    },
    {
      "epoch": 0.8791866028708134,
      "grad_norm": 0.3639909029006958,
      "learning_rate": 1.9401107929041795e-05,
      "loss": 0.7291,
      "step": 10290
    },
    {
      "epoch": 0.880041011619959,
      "grad_norm": 0.3874698281288147,
      "learning_rate": 1.9291749867913225e-05,
      "loss": 0.6968,
      "step": 10300
    },
    {
      "epoch": 0.8808954203691046,
      "grad_norm": 0.39110490679740906,
      "learning_rate": 1.91735909707114e-05,
      "loss": 0.7166,
      "step": 10310
    },
    {
      "epoch": 0.8817498291182502,
      "grad_norm": 0.39673879742622375,
      "learning_rate": 1.9046743759101545e-05,
      "loss": 0.6933,
      "step": 10320
    },
    {
      "epoch": 0.8826042378673957,
      "grad_norm": 0.4197354316711426,
      "learning_rate": 1.8911329028553324e-05,
      "loss": 0.7225,
      "step": 10330
    },
    {
      "epoch": 0.8834586466165414,
      "grad_norm": 0.3761933147907257,
      "learning_rate": 1.876747573330835e-05,
      "loss": 0.7163,
      "step": 10340
    },
    {
      "epoch": 0.8843130553656869,
      "grad_norm": 0.36793383955955505,
      "learning_rate": 1.8615320863578244e-05,
      "loss": 0.7081,
      "step": 10350
    },
    {
      "epoch": 0.8851674641148325,
      "grad_norm": 0.4014787971973419,
      "learning_rate": 1.8455009315090036e-05,
      "loss": 0.7136,
      "step": 10360
    },
    {
      "epoch": 0.8860218728639782,
      "grad_norm": 0.35117441415786743,
      "learning_rate": 1.8286693751103294e-05,
      "loss": 0.7093,
      "step": 10370
    },
    {
      "epoch": 0.8868762816131237,
      "grad_norm": 0.39133623242378235,
      "learning_rate": 1.811053445703024e-05,
      "loss": 0.7399,
      "step": 10380
    },
    {
      "epoch": 0.8877306903622693,
      "grad_norm": 0.3980655074119568,
      "learning_rate": 1.7926699187797375e-05,
      "loss": 0.7088,
      "step": 10390
    },
    {
      "epoch": 0.8885850991114149,
      "grad_norm": 0.4149200916290283,
      "learning_rate": 1.7735363008093995e-05,
      "loss": 0.7443,
      "step": 10400
    },
    {
      "epoch": 0.8894395078605605,
      "grad_norm": 0.3846627175807953,
      "learning_rate": 1.7536708125659577e-05,
      "loss": 0.7065,
      "step": 10410
    },
    {
      "epoch": 0.8902939166097061,
      "grad_norm": 0.4279932975769043,
      "learning_rate": 1.733092371776905e-05,
      "loss": 0.7121,
      "step": 10420
    },
    {
      "epoch": 0.8911483253588517,
      "grad_norm": 0.36222484707832336,
      "learning_rate": 1.7118205751080915e-05,
      "loss": 0.7117,
      "step": 10430
    },
    {
      "epoch": 0.8920027341079972,
      "grad_norm": 0.3647192120552063,
      "learning_rate": 1.6898756795019935e-05,
      "loss": 0.7503,
      "step": 10440
    },
    {
      "epoch": 0.8928571428571429,
      "grad_norm": 0.367274671792984,
      "learning_rate": 1.667278582887211e-05,
      "loss": 0.7105,
      "step": 10450
    },
    {
      "epoch": 0.8937115516062885,
      "grad_norm": 0.3640269637107849,
      "learning_rate": 1.6440508042775545e-05,
      "loss": 0.7023,
      "step": 10460
    },
    {
      "epoch": 0.894565960355434,
      "grad_norm": 0.38335949182510376,
      "learning_rate": 1.6202144632796796e-05,
      "loss": 0.7173,
      "step": 10470
    },
    {
      "epoch": 0.8954203691045797,
      "grad_norm": 0.3835354745388031,
      "learning_rate": 1.595792259028783e-05,
      "loss": 0.731,
      "step": 10480
    },
    {
      "epoch": 0.8962747778537252,
      "grad_norm": 0.3645132780075073,
      "learning_rate": 1.570807448572419e-05,
      "loss": 0.6889,
      "step": 10490
    },
    {
      "epoch": 0.8971291866028708,
      "grad_norm": 0.37407055497169495,
      "learning_rate": 1.5452838247230197e-05,
      "loss": 0.737,
      "step": 10500
    },
    {
      "epoch": 0.8979835953520164,
      "grad_norm": 0.3990010619163513,
      "learning_rate": 1.5192456934002106e-05,
      "loss": 0.7042,
      "step": 10510
    },
    {
      "epoch": 0.898838004101162,
      "grad_norm": 0.3691908121109009,
      "learning_rate": 1.492717850484502e-05,
      "loss": 0.706,
      "step": 10520
    },
    {
      "epoch": 0.8996924128503075,
      "grad_norm": 0.37118005752563477,
      "learning_rate": 1.4657255582043904e-05,
      "loss": 0.7115,
      "step": 10530
    },
    {
      "epoch": 0.9005468215994532,
      "grad_norm": 0.3894721567630768,
      "learning_rate": 1.4382945210793663e-05,
      "loss": 0.703,
      "step": 10540
    },
    {
      "epoch": 0.9014012303485988,
      "grad_norm": 0.39006462693214417,
      "learning_rate": 1.4104508614417267e-05,
      "loss": 0.7172,
      "step": 10550
    },
    {
      "epoch": 0.9022556390977443,
      "grad_norm": 0.368083655834198,
      "learning_rate": 1.382221094560512e-05,
      "loss": 0.7085,
      "step": 10560
    },
    {
      "epoch": 0.90311004784689,
      "grad_norm": 0.3643242120742798,
      "learning_rate": 1.3536321033912506e-05,
      "loss": 0.7186,
      "step": 10570
    },
    {
      "epoch": 0.9039644565960355,
      "grad_norm": 0.3984137177467346,
      "learning_rate": 1.3247111129755609e-05,
      "loss": 0.7063,
      "step": 10580
    },
    {
      "epoch": 0.9048188653451812,
      "grad_norm": 0.43620651960372925,
      "learning_rate": 1.2954856645149868e-05,
      "loss": 0.7116,
      "step": 10590
    },
    {
      "epoch": 0.9056732740943267,
      "grad_norm": 0.3951973021030426,
      "learning_rate": 1.2659835891437553e-05,
      "loss": 0.7048,
      "step": 10600
    },
    {
      "epoch": 0.9065276828434723,
      "grad_norm": 0.4190177619457245,
      "learning_rate": 1.236232981425436e-05,
      "loss": 0.744,
      "step": 10610
    },
    {
      "epoch": 0.907382091592618,
      "grad_norm": 0.38798749446868896,
      "learning_rate": 1.2062621725987402e-05,
      "loss": 0.707,
      "step": 10620
    },
    {
      "epoch": 0.9082365003417635,
      "grad_norm": 0.3879949450492859,
      "learning_rate": 1.1760997035979318e-05,
      "loss": 0.6804,
      "step": 10630
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 0.3721603453159332,
      "learning_rate": 1.1457742978735514e-05,
      "loss": 0.7295,
      "step": 10640
    },
    {
      "epoch": 0.9099453178400547,
      "grad_norm": 0.3837917447090149,
      "learning_rate": 1.1153148340393298e-05,
      "loss": 0.6877,
      "step": 10650
    },
    {
      "epoch": 0.9107997265892003,
      "grad_norm": 0.3692246079444885,
      "learning_rate": 1.0847503183713424e-05,
      "loss": 0.6994,
      "step": 10660
    },
    {
      "epoch": 0.9116541353383458,
      "grad_norm": 0.3923785984516144,
      "learning_rate": 1.0541098571855896e-05,
      "loss": 0.7272,
      "step": 10670
    },
    {
      "epoch": 0.9125085440874915,
      "grad_norm": 0.3547303080558777,
      "learning_rate": 1.0234226291203123e-05,
      "loss": 0.6856,
      "step": 10680
    },
    {
      "epoch": 0.913362952836637,
      "grad_norm": 0.36296340823173523,
      "learning_rate": 9.927178573494333e-06,
      "loss": 0.6873,
      "step": 10690
    },
    {
      "epoch": 0.9142173615857826,
      "grad_norm": 0.4047708213329315,
      "learning_rate": 9.620247817535902e-06,
      "loss": 0.7264,
      "step": 10700
    },
    {
      "epoch": 0.9150717703349283,
      "grad_norm": 0.3722198009490967,
      "learning_rate": 9.313726310752581e-06,
      "loss": 0.7038,
      "step": 10710
    },
    {
      "epoch": 0.9159261790840738,
      "grad_norm": 0.3617464005947113,
      "learning_rate": 9.00790595084477e-06,
      "loss": 0.6975,
      "step": 10720
    },
    {
      "epoch": 0.9167805878332194,
      "grad_norm": 0.40016502141952515,
      "learning_rate": 8.703077967816944e-06,
      "loss": 0.7111,
      "step": 10730
    },
    {
      "epoch": 0.917634996582365,
      "grad_norm": 0.40464913845062256,
      "learning_rate": 8.39953264664192e-06,
      "loss": 0.7216,
      "step": 10740
    },
    {
      "epoch": 0.9184894053315106,
      "grad_norm": 0.36447587609291077,
      "learning_rate": 8.097559050825058e-06,
      "loss": 0.6842,
      "step": 10750
    },
    {
      "epoch": 0.9193438140806562,
      "grad_norm": 0.39392754435539246,
      "learning_rate": 7.79744474713166e-06,
      "loss": 0.7166,
      "step": 10760
    },
    {
      "epoch": 0.9201982228298018,
      "grad_norm": 0.35473698377609253,
      "learning_rate": 7.499475531739696e-06,
      "loss": 0.6813,
      "step": 10770
    },
    {
      "epoch": 0.9210526315789473,
      "grad_norm": 0.3602163791656494,
      "learning_rate": 7.203935158078662e-06,
      "loss": 0.6813,
      "step": 10780
    },
    {
      "epoch": 0.921907040328093,
      "grad_norm": 0.36335763335227966,
      "learning_rate": 6.911105066613709e-06,
      "loss": 0.6888,
      "step": 10790
    },
    {
      "epoch": 0.9227614490772386,
      "grad_norm": 0.38769182562828064,
      "learning_rate": 6.621264116832394e-06,
      "loss": 0.6822,
      "step": 10800
    },
    {
      "epoch": 0.9236158578263841,
      "grad_norm": 0.3771034777164459,
      "learning_rate": 6.334688321689274e-06,
      "loss": 0.6823,
      "step": 10810
    },
    {
      "epoch": 0.9244702665755298,
      "grad_norm": 0.41323184967041016,
      "learning_rate": 6.051650584761237e-06,
      "loss": 0.6902,
      "step": 10820
    },
    {
      "epoch": 0.9253246753246753,
      "grad_norm": 0.37754902243614197,
      "learning_rate": 5.772420440363841e-06,
      "loss": 0.6825,
      "step": 10830
    },
    {
      "epoch": 0.9261790840738209,
      "grad_norm": 0.37076061964035034,
      "learning_rate": 5.497263796876207e-06,
      "loss": 0.674,
      "step": 10840
    },
    {
      "epoch": 0.9270334928229665,
      "grad_norm": 0.3934485912322998,
      "learning_rate": 5.226442683518813e-06,
      "loss": 0.6977,
      "step": 10850
    },
    {
      "epoch": 0.9278879015721121,
      "grad_norm": 0.39820343255996704,
      "learning_rate": 4.960215000825432e-06,
      "loss": 0.6989,
      "step": 10860
    },
    {
      "epoch": 0.9287423103212576,
      "grad_norm": 0.38825860619544983,
      "learning_rate": 4.698834275046711e-06,
      "loss": 0.7128,
      "step": 10870
    },
    {
      "epoch": 0.9295967190704033,
      "grad_norm": 0.3672446012496948,
      "learning_rate": 4.442549416719423e-06,
      "loss": 0.6864,
      "step": 10880
    },
    {
      "epoch": 0.9304511278195489,
      "grad_norm": 0.37998348474502563,
      "learning_rate": 4.1916044836311565e-06,
      "loss": 0.7349,
      "step": 10890
    },
    {
      "epoch": 0.9313055365686944,
      "grad_norm": 0.3763866424560547,
      "learning_rate": 3.94623844840626e-06,
      "loss": 0.6826,
      "step": 10900
    },
    {
      "epoch": 0.9321599453178401,
      "grad_norm": 0.38429591059684753,
      "learning_rate": 3.7066849709343467e-06,
      "loss": 0.7069,
      "step": 10910
    },
    {
      "epoch": 0.9330143540669856,
      "grad_norm": 0.3733646869659424,
      "learning_rate": 3.473172175858059e-06,
      "loss": 0.6772,
      "step": 10920
    },
    {
      "epoch": 0.9338687628161313,
      "grad_norm": 0.3572109341621399,
      "learning_rate": 3.2459224353320046e-06,
      "loss": 0.6732,
      "step": 10930
    },
    {
      "epoch": 0.9347231715652768,
      "grad_norm": 0.3500171899795532,
      "learning_rate": 3.025152157259735e-06,
      "loss": 0.7086,
      "step": 10940
    },
    {
      "epoch": 0.9355775803144224,
      "grad_norm": 0.3593735098838806,
      "learning_rate": 2.811071579210425e-06,
      "loss": 0.6761,
      "step": 10950
    },
    {
      "epoch": 0.9364319890635681,
      "grad_norm": 0.3791474997997284,
      "learning_rate": 2.603884568211513e-06,
      "loss": 0.6875,
      "step": 10960
    },
    {
      "epoch": 0.9372863978127136,
      "grad_norm": 0.3667813241481781,
      "learning_rate": 2.4037884266079434e-06,
      "loss": 0.6905,
      "step": 10970
    },
    {
      "epoch": 0.9381408065618592,
      "grad_norm": 0.3938116431236267,
      "learning_rate": 2.2109737041728955e-06,
      "loss": 0.7121,
      "step": 10980
    },
    {
      "epoch": 0.9389952153110048,
      "grad_norm": 0.3863193392753601,
      "learning_rate": 2.0256240166489454e-06,
      "loss": 0.7173,
      "step": 10990
    },
    {
      "epoch": 0.9398496240601504,
      "grad_norm": 0.3599783778190613,
      "learning_rate": 1.847915870892413e-06,
      "loss": 0.7121,
      "step": 11000
    },
    {
      "epoch": 0.9407040328092959,
      "grad_norm": 0.34297803044319153,
      "learning_rate": 1.6780184967874715e-06,
      "loss": 0.6833,
      "step": 11010
    },
    {
      "epoch": 0.9415584415584416,
      "grad_norm": 0.3635624051094055,
      "learning_rate": 1.5160936860900294e-06,
      "loss": 0.6911,
      "step": 11020
    },
    {
      "epoch": 0.9424128503075871,
      "grad_norm": 0.36491236090660095,
      "learning_rate": 1.3622956383548762e-06,
      "loss": 0.6781,
      "step": 11030
    },
    {
      "epoch": 0.9432672590567327,
      "grad_norm": 0.3706292510032654,
      "learning_rate": 1.216770814092834e-06,
      "loss": 0.6942,
      "step": 11040
    },
    {
      "epoch": 0.9441216678058784,
      "grad_norm": 0.379960834980011,
      "learning_rate": 1.0796577952977013e-06,
      "loss": 0.6904,
      "step": 11050
    },
    {
      "epoch": 0.9449760765550239,
      "grad_norm": 0.36443254351615906,
      "learning_rate": 9.510871534758705e-07,
      "loss": 0.6588,
      "step": 11060
    },
    {
      "epoch": 0.9458304853041695,
      "grad_norm": 0.4932965636253357,
      "learning_rate": 8.311813253042334e-07,
      "loss": 0.7023,
      "step": 11070
    },
    {
      "epoch": 0.9466848940533151,
      "grad_norm": 0.3769393563270569,
      "learning_rate": 7.200544960348308e-07,
      "loss": 0.6649,
      "step": 11080
    },
    {
      "epoch": 0.9475393028024607,
      "grad_norm": 0.3323296308517456,
      "learning_rate": 6.178124907572545e-07,
      "loss": 0.6689,
      "step": 11090
    },
    {
      "epoch": 0.9483937115516062,
      "grad_norm": 0.3671296238899231,
      "learning_rate": 5.245526736223528e-07,
      "loss": 0.7066,
      "step": 11100
    },
    {
      "epoch": 0.9492481203007519,
      "grad_norm": 0.35006770491600037,
      "learning_rate": 4.403638551232205e-07,
      "loss": 0.6818,
      "step": 11110
    },
    {
      "epoch": 0.9501025290498974,
      "grad_norm": 0.4351028501987457,
      "learning_rate": 3.6532620752175616e-07,
      "loss": 0.6931,
      "step": 11120
    },
    {
      "epoch": 0.9509569377990431,
      "grad_norm": 0.3864075243473053,
      "learning_rate": 2.9951118850133247e-07,
      "loss": 0.694,
      "step": 11130
    },
    {
      "epoch": 0.9518113465481887,
      "grad_norm": 0.3760283291339874,
      "learning_rate": 2.429814731182774e-07,
      "loss": 0.69,
      "step": 11140
    },
    {
      "epoch": 0.9526657552973342,
      "grad_norm": 0.37165361642837524,
      "learning_rate": 1.957908941169813e-07,
      "loss": 0.6908,
      "step": 11150
    },
    {
      "epoch": 0.9535201640464799,
      "grad_norm": 0.46264365315437317,
      "learning_rate": 1.5798439066545254e-07,
      "loss": 0.7114,
      "step": 11160
    },
    {
      "epoch": 0.9543745727956254,
      "grad_norm": 0.35400986671447754,
      "learning_rate": 1.2959796556014471e-07,
      "loss": 0.7024,
      "step": 11170
    },
    {
      "epoch": 0.955228981544771,
      "grad_norm": 0.3735068440437317,
      "learning_rate": 1.1065865094081884e-07,
      "loss": 0.698,
      "step": 11180
    },
    {
      "epoch": 0.9560833902939166,
      "grad_norm": 0.36363786458969116,
      "learning_rate": 1.011844825480734e-07,
      "loss": 0.6823,
      "step": 11190
    },
    {
      "epoch": 0.9564251537935748,
      "step": 11194,
      "total_flos": 4.547514943786923e+19,
      "train_loss": 0.0040640543247091356,
      "train_runtime": 609.7761,
      "train_samples_per_second": 220.291,
      "train_steps_per_second": 18.358
    }
  ],
  "logging_steps": 10,
  "max_steps": 11194,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 159,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 4.547514943786923e+19,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}